Skip to main content

WebCortex

pip install web-cortex-framework

Installs as web-cortex-framework, imports as webcortex — the same split as djangorestframework → import rest_framework.

Python 3.12, 3.13 and 3.14, including the free-threaded build (3.14t), which is the fast path for Python-backed routes.

📖 Documentation · Tutorial · Orchestration · Token economy · Security

A Python web framework with a Rust core, built on one idea:

If you declared it, Rust can run it — and an agent can call it.

Django and Rails were designed when the only client was a browser and the only author was a person. Today the client is just as likely to be a model, and so is the author. WebCortex treats both as the primary case: every declaration is simultaneously a REST route, an OpenAPI operation and an MCP tool; agents, behaviours and flows are first-class and compose; and the whole application can describe itself to the model that is writing it.

# api.py
from webcortex import WebCortex

app = WebCortex("bookstore", database="sqlite://./app.db")

app.api_key("WEBCORTEX_API_KEY", id="service", scopes=["read", "write"])
app.rate_limit(per_second=50)
app.anonymous_scopes("read")

app.resource(
    "books",
    fields={"id": int, "title": str, "author": str, "year": int},
    tools=True,
    read_scopes=["read"],
    write_scopes=["write"],
)

@app.get("/books/{id}/blurb", tool=True, scopes=["read"])
def blurb(id: int) -> str:
    """One-line pitch for a book."""
    return f"Book {id} — highly recommended."
$ webcortex dev

You now have a REST API, an OpenAPI 3.1 document, a live MCP server exposing all six endpoints as tools, authentication, rate limiting, and security headers. No second file, no schema written twice, no drift.

Twenty more lines make it a multi-agent system:

app.models(default="claude-opus-5", fast="ollama/qwen3.5:9b")   # tiers, not models

app.context("policy", data={"max_discount": 0.2})              # what agents know at step one
notes = app.memory("notes", scopes=["read"])                     # a durable, per-user scratchpad

app.agent("librarian", description="Finds and pitches books.",
          tools=["list_books", "get_books_by_id_blurb"], context=["policy"],
          memory="notes", scopes=["read"], token_budget=60_000)

app.agent("front_desk", handoffs=["librarian"], scopes=["read"],
          context_window=40_000, token_budget=60_000,
          expose_at="/ask")                                     # hands off; compacts

app.flow("pitch_all", parallel=["librarian", "librarian"],
         token_budget=120_000)                                  # orchestration as data

# anonymous_scopes("read") above means anyone, with no key, may start /ask,
# /agents/librarian and /flows/pitch-all: the token budgets and the rate limit
# are what bound that. Use expose_scopes=["write"] to keep a run for key holders.

Every one of those is a route, a tool, and an MCP entry. The front desk hands a conversation to the librarian; the librarian remembers what it learned about the caller; the whole run, handoff included, is bounded by one budget; and webcortex context prints the lot in a form a coding model can extend.


Start here

uv venv --python 3.13 && source .venv/bin/activate   # puts `webcortex` on PATH
uv pip install web-cortex-framework

webcortex new myapp                # --template api | fullstack | agent | behaviour | orchestration | robotics | hub
cd myapp
export WEBCORTEX_API_KEY=$(webcortex keygen)
webcortex doctor                   # anything else to set first: other keys, model provider keys
webcortex dev

Why a Rust core, specifically

Most Rust-accelerated Python servers put Rust at the socket and call Python for every request. You get faster parsing; your handler is still interpreted.

WebCortex puts the boundary somewhere more useful. Python is a declaration language that compiles to a plan the Rust runtime executes. A route whose work is expressible as data — a query, a proxy, a rendered page, a static file, an agent invocation, a flow — runs entirely in Rust and never enters the interpreter at request time. In practice that is most of a CRUD API and all of an orchestration.

$ webcortex dev                # the boot banner
  19 routes, 17 served without touching Python

When a route genuinely needs Python, it crosses onto a pool of free-threaded interpreter workers (free-threaded CPython 3.14, GIL disabled), each running its own event loop. Handlers run in real parallel — measured at 4.82× vs 1.38× under the GIL (DESIGN.md has the numbers and their caveats).

WebCortex runs correctly on a GIL build too, and tells you which mode it is in.

Behaviours

A "skill" written as a prompt is a suggestion. The model reads it and may ignore it, and "if X then Y" fails silently when it does.

A Behaviour inverts that. The control flow is real Python — a for loop is a loop, an if is a branch, and both execute whether or not a model would have chosen to. Only the leaves are probabilistic, and in v2 the leaves run concurrently:

@app.behaviour("triage", tools=["list_tickets", "update_tickets"],
               scopes=["read", "write"],                       # unscoped = anyone may start it
               model="fast", max_steps=200, token_budget=100_000)
def triage(ctx, input):
    """Classify every open ticket at once and escalate the urgent ones."""
    tickets = [t for t in ctx.call("list_tickets", limit=50) if t["state"] == "open"]

    verdicts = ctx.ask_many(                                   # fifty model calls, one wait
        [f"Grade this ticket:\n{t['body']}" for t in tickets],
        schema={"type": "object",
                "properties": {"urgency": {"type": "integer"},
                               "category": {"enum": ["bug", "billing", "other"]}},
                "required": ["urgency", "category"]},
    )

    escalated = [t["id"] for t, v in zip(tickets, verdicts)     # a real branch
                 if v["urgency"] >= input.get("threshold", 7)]

    ctx.gather(*[                                                # fifty writes, one wait
        ("update_tickets", {**t, "urgency": v["urgency"], "category": v["category"],
                            "state": "escalated" if t["id"] in escalated else "triaged"})
        for t, v in zip(tickets, verdicts)
    ])
    return {"escalated": escalated, "usage": ctx.usage}

You get a procedure with deterministic structure and probabilistic steps, rather than a probabilistic procedure.

ctx is how a behaviour reaches the world:

ctx.call(tool, **kwargs) invoke one of the app's tools, in-process
ctx.gather((tool, kwargs), ...) several tool calls, concurrently
ctx.ask(prompt, schema=..., model="fast") a model call; a schema forces the shape; the model may be a tier
ctx.ask_many([prompts], schema=...) the classification loop collapsed into one wait
ctx.context(name) a declared context provider, resolved on demand
ctx.log(msg) / ctx.halt(reason) narrate or stop deliberately
ctx.usage / ctx.trace budget consumed (this run and the whole tree) and every leaf executed
ctx.user / ctx.tools the delegated principal and what it may call

A behaviour is a tool. It registers as a route, so it is automatically an MCP tool, an OpenAPI operation, and something an agent — or another behaviour, or a flow — can call.

The runtime enforces the limits, not your diligence: max_steps caps leaf operations; token_budget caps spend for the run and everything it calls; a behaviour cannot call a tool it did not declare; an approval-gated tool cannot be laundered through a behaviour or a gather; scopes are delegated by intersection, never unioned.

Orchestration

One agent is a tool loop. Several are a system, and a system needs answers a loop never asks: who is in control, what may it do, how much may the whole thing cost, what happens when a human has to decide. Four primitives, all executed by the runtime:

Every agent is a tool. An agent is mounted at /agents/<name> (underscores become hyphens) and exposed under its own name, so app.agent("editor", tools=["researcher", "writer"]) is the entire supervisor/worker pattern. Workers run under a principal delegated from the supervisor's, one nesting level deeper, against the supervisor's budget.

Handoffs. app.agent("front_desk", handoffs=["billing", "technical"]) gives the front desk transfer_to_* tools. Calling one moves the conversation to the specialist — its system prompt, tools and context apply from the next step — while the budget and the caller's authority carry over. Authority can only shrink along a chain.

Flows: orchestration as data.

app.flow("briefing", pipeline=["researcher", "writer"], token_budget=150_000)
app.flow("review",   parallel=["security_review", "style_review"], merge="collect")
app.flow("desk",     route={"billing": "billing", "technical": "technical"},
                     default="front_desk", classify_with="fast")

Every step is a tool — an agent, a behaviour, another flow, a plain route — so composition is uniform, and a flow is itself a tool. Steps map arguments with {"tool": "x", "input": {"id": "$.id", "q": "$input.query"}}.

Sessions. Post {"input": "...", "session_id": "..."} and the conversation continues; the key includes the principal, so callers never see each other's history, and anonymous callers, who share one identity, cannot open one. Long sessions are compacted, not truncated.

Approvals that resume. A gated tool suspends the run; a human decides at POST /_webcortex/approvals/{id} with {"approve": true|false, "note": "..."}; the run continues — including the rest of the turn it was interrupted in. A denial is a tool error the model reads and reacts to.

Budgets compose. The outermost run's token_budget is shared by every agent, behaviour and flow it calls. usage.tree_tokens reports the total.

Token economy

The cost of an agent system is which model answers, how much context each call carries, and how many calls are made. Each has a declaration.

app.models(default="claude-opus-5", fast="claude-haiku-4-5-20251001",
           local="ollama/qwen3.5:9b")
app.provider("groq", base_url="https://api.groq.com/openai/v1", api_key_env="GROQ_API_KEY")
app.pricing("claude-opus-5", input_per_mtok=5, output_per_mtok=25)
  • Tiers, not models. Judgement uses default; classification, extraction and routing use fast. Moving a workload to a cheaper or local model is one edit.
  • Two wire formats, chosen by prefix. Anthropic Messages and OpenAI Chat Completions — which is what Ollama, vLLM, LM Studio, Groq and OpenRouter all speak. ollama/qwen3.5:9b needs no key. Deliberately not a universal LLM abstraction.
  • Prompt caching on by default: the system prompt, context and tool definitions are cached across the steps of a run.
  • Bounded context. tool_result_limit caps what the model sees of any tool result; context_window compacts older turns with the fast model when the measured input exceeds it.
  • A ledger. GET /_webcortex/usage reports tokens by agent, behaviour, flow and model, and dollars where you declared prices — null, not zero, where you did not.

Evidence: typed answers, traces, evals

New in 2.8: what an agent answered, what a run did, and whether it still does it.

@dataclass
class Triage:
    category: Literal["bug", "billing", "other"]
    urgency: int

app.agent("triage", system="Classify the ticket.", output=Triage)
app.agent("local", model="ollama/qwen3.5:9b", options={"seed": 7, "keep_alive": "30m"})

app.eval("refund_goes_to_billing", target="front_desk", input="I was charged twice",
         expect={"path": ["front_desk", "billing"], "tools_called": ["lookup_invoice"],
                 "max_tokens": 6000})
  • Structured output. output= is a dataclass, a TypedDict or a JSON Schema. The agent answers through a final_answer tool and the runtime validates it: a mismatch goes back to the model with the list of problems, twice, then the run fails with WC318. The result carries the object as data, and a flow hands it to the next step as an object. docs
  • Request options. options= passes provider-specific fields straight into the request body (seed, top_p, keep_alive, thinking, …) per agent, per behaviour or per provider, without a keyword for each. None removes a field. Still not a universal LLM abstraction: nothing is interpreted or renamed. docs
  • Traces. Every agent, flow and behaviour run is a span tree: the run, each model call with its tokens, each tool call, nested runs under the call that started them. webcortex trace, GET /_webcortex/traces, the webcortex_trace MCP tool, and OTLP/HTTP export to the collector you already run. An incoming W3C traceparent is joined. Spans hold names, timings, token counts and error codes, never prompts or results. docs
  • Evals with cassettes. webcortex eval --record keeps the model's responses; afterwards evals replay them in-process with no key and no tokens, while your tools, scopes, handoffs, budgets and flows run for real. Change an agent's prompt, model, tools or options and its recording goes stale (WC319) instead of passing on old answers. docs
$ webcortex trace latest
  trace 7d1c…e02a  ok  2.41s  4 model call(s)  2 tool call(s)  1830 tokens
  agent front_desk  2.41s  completed
    └ chat default  0.61s  412 in / 38 out
    └ tool billing  1.52s
      └ agent billing  1.52s  completed
        └ chat default  0.58s  390 in / 41 out
        └ tool lookup_invoice  3ms
        └ chat default  0.90s  455 in / 96 out
    └ chat default  0.27s  362 in / 36 out

(The tree's shape is what the command prints; the timings and token counts above are an illustration, not a measurement.)

Context and memory

app.context("policy", data={"refund_days": 30})
app.context("open_queue", sql="SELECT id, kind FROM tickets WHERE state='open' LIMIT 20")
app.context("my_orders", sql="SELECT * FROM orders WHERE customer = ?", params=["@principal"])

@app.context("account")
def account(req) -> dict: ...

notes = app.memory("notes", read_scopes=["read"], write_scopes=["write"])
app.agent("assistant", context=["policy", "my_orders"], memory="notes", ...)

A context provider is resolved when a run starts and injected into the system prompt, bounded by max_chars. @principal binds to the human behind however many agents deep the call is. A memory is four Rust-executed tools — remember, recall, search, forget — keyed by that same principal, so an agent writing on someone's behalf writes to that someone's memory and can never read another's.

The nine kinds of route

Kind Declared with Runs in
Static app.static(...) Rust
Query app.query(...), app.resource(...), app.memory(...) Rust
Page app.page(...) Rust (minijinja)
Files app.static_files(...) Rust
Proxy app.proxy(...) Rust
Agent app.agent(...) Rust
Flow app.flow(...) Rust
Behaviour @app.behaviour(...) Python worker pool
Python @app.get(...) Python worker pool

Every route is a tool

Mark a route tool=True and it appears in MCP tools/list, with an input schema derived from the handler's own type hints. Agents declared in the same app call those tools in-process — a function call through the same dispatcher the HTTP server uses, not a loopback request. A typo in tools, handoffs or context fails at boot with a "did you mean" suggestion.

What makes agents safe to deploy

Enforced by the runtime, not by your diligence:

Delegated authority. A run executes as caller.delegate_to_agent(...), whose scopes are intersected with the caller's — never unioned — and a handoff intersects again. An anonymous caller cannot launch a privileged agent.

Human approval gates. Mark a route approval="required" and an agent asking for it does not get it. An agent's run suspends until a human decides, then resumes with the rest of its turn. Every other path a model could take is refused: a direct MCP call is rejected, a behaviour (alone or in gather) halts, and a flow that contains a gated tool does not boot. Resuming continues the agent that hit the gate, not a supervisor that called it, so give gated tools to the agents people call directly. The gate governs models, not people: a caller holding the route's scopes can still call it over plain HTTP.

Runtime-enforced budgets that compose. max_steps per run; token_budget for the run and everything under it.

Scope-filtered tool lists. tools/list shows only what that caller can invoke.

A full audit trail, including refused calls, handoffs, compactions and approval decisions, at GET /_webcortex/audit.

Agents that see, machines that move

WebCortex embeds no vision model — it carries pixels to the model that has one, and puts the guard rails a physical system needs around the tools that move it.

from webcortex import HTTPError, Image, WebCortex

app = WebCortex("cell", start_halted=True)            # boots with the e-stop engaged

@app.get("/camera", tool=True, tool_name="camera", scopes=["observe"])
def camera() -> dict:
    frame = cam.read()                                 # an OpenCV / NumPy frame
    return {"frame": Image.from_array(frame, bgr=True), "exposure_ms": 8}

@app.post("/arm/move", tool=True, tool_name="move_arm", scopes=["operate"],
          actuator=True, approval="required")
def move(joint: str, degrees: float) -> dict:
    if not -90 <= degrees <= 90:
        raise HTTPError(422, "out of range")           # limits live in code, not prompts
    return arm.move(joint, degrees)

app.agent("operator", tools=["camera", "move_arm"], scopes=["observe", "operate"], max_images=3)
  • The agent sees the frame. A tool that returns a webcortex.Image — anywhere in its result — hands the model the picture itself, over both wire formats (Anthropic and OpenAI-compatible, so local vision models too) and over MCP, where Claude Code and other clients receive it as image content. Agent endpoints also take "images": [...] with the input.
  • Frames do not become a token leak. Images are lifted out before tool-result bounding, step records keep their size rather than their base64, and only the last max_images stay in context as pixels.
  • An emergency stop that cannot be routed around. actuator=True routes refuse with 423 while halted — over HTTP, from an agent, a behaviour, a flow, MCP, or an approval granted after the stop. webcortex halt / webcortex release, or POST /_webcortex/halt and /release.
  • webcortex new cell -t robotics scaffolds a simulated arm and camera with an inspector agent and an operator agent, gated, scoped and booting halted. Swap the Cell class for your driver — serial, ROS 2, Modbus, a vendor SDK.

See Vision and robotics.

The device hub: cameras in, insights out

WebCortex sits between devices and agents, and the Rust core does the work. Cameras and sensors stream in, agents see what they send, and what the agents conclude streams out to anyone subscribed.

app.camera("dock", max_fps=10)                                   # pushed: browser, Pi, ESP32, curl
app.camera("lobby", source="http://192.168.1.40/snapshot.jpg")   # pulled: any IP camera
app.sensor("dock_env")                                           # telemetry only

app.agent("inspector", tools=["dock_snapshot", "dock_env_telemetry"], max_images=2)
app.watch("dock_watch", device="dock", agent="inspector", every=15,
          input="Anything unsafe at the dock?", max_runs_per_hour=120,
          webhook="https://ops.example.com/hooks/dock")
  • Connect anything. WebSocket (/devices/dock/ws: binary messages are frames, text is telemetry), HTTP POST, or an IP camera's snapshot URL. Open /devices/dock/connect on a phone or laptop and its camera becomes the device.
  • No Python per frame. Ring buffers, sniffing, per-device rate limits and broadcast fan-out all live in Rust.
  • Agents see it. dock_snapshot is a tool that returns the live frame as an image the model sees, over MCP too.
  • Watchers are the middle of the hub. When the picture changes, the agent looks, and its insight goes to every /devices/dock/stream subscriber, to a webhook, and to the audit log. Spend is capped per hour.
  • Agents over WebSocket. Any agent route accepts a socket: send a turn, receive each step live, then the result.
  • A stdlib Python client (webcortex.client): DeviceConnection for drivers, subscribe() for consumers, AgentSocket for services.

webcortex new site -t hub scaffolds it. See The device hub.

Security defaults

Deny-by-default throughout; relaxing something costs a line, tightening it costs nothing. API keys are referenced by environment variable, hashed with SHA-256, and compared in constant time. JWT (HS, RS, ES and EdDSA) with mandatory expiry, and audience and issuer enforced whenever they are configured. Per-principal token-bucket rate limiting. Security headers on every response. CORS that refuses * with credentials at boot. Path traversal, symlink escapes, and dotfiles refused by the static server. The control plane — including approvals, usage and models — requires webcortex:admin once any authentication is configured.

The frontend, without the mess

Two clean paths sharing one data layer, chosen per route: server-rendered pages executed in Rust (app.page("/", "index.html", sql=..., bind="books")), where a template receives a data object and nothing else; and a typed TypeScript client for SPA frontends, generated from the same route table by webcortex typegen.

AI-native development

webcortex new writes an AGENTS.md (read by Codex, Cursor and most agents) and a CLAUDE.md that imports it, so the rules travel with the project; webcortex context describes the running app for the model extending it; webcortex mcp-config prints the one command that plugs the app into Claude Code, Claude Desktop or Cursor.

$ webcortex context                     # the app, described for a coding model
$ webcortex evolve "add reviews tied to books and a behaviour that summarises them" --model fast
$ webcortex check                       # boot-time validation, with hints
$ webcortex doctor                      # keys present? models resolve? port free? e-stop?
$ webcortex status                      # what the running app is doing, and what just failed
$ webcortex explain WC402               # what an error code means and what to do

Every failure carries a stable code and a hint — in HTTP error bodies, agent runs, MCP tool errors and the CLI — and the live view at /_webcortex/status (also the webcortex_status MCP tool) shows recent failures by code and request id, runs in flight, spend and devices. On eighteen scripted broken scenarios, 2.5.0 gives a code and a hint for 18/18 where 2.4.0 gave none; see benchmarks/diagnostics/RESULTS.md and docs/diagnostics.md.

The context pack is the app's shape — routes, tools and schemas, agents, behaviours, flows, context, memory, models, security posture — derived from the manifest in a few thousand tokens, plus a cheat sheet of the framework's API. evolve feeds it to a model (with the app's own aliases, so --model fast can be a local Ollama model) and prints a proposal to review. The loop is describe → propose → check → run, and boot-time validation is what makes it safe to repeat.

AGENTS.md is the machine-facing reference for tools that edit this repository or write apps on it. scout/ is a local-model reviewer that leaves suggestions for the framework's own next iteration.

Commands

webcortex new <name>      scaffold a project (api | fullstack | agent | behaviour | orchestration | robotics | hub)
webcortex dev             run with a startup report
webcortex check           validate; the route table, tools, agents, flows and public surface
webcortex security        what is reachable without a credential
webcortex tools           the agent tool manifest
webcortex context         the context pack, for an AI coding tool
webcortex mcp-config      the MCP client config and `claude mcp add` command for this app
webcortex evolve "…"      ask a model to propose an extension
webcortex typegen         generate a typed TypeScript client
webcortex openapi         the OpenAPI 3.1 document
webcortex sql             DDL for declared resources and memories
webcortex keygen          mint an API key
webcortex doctor          check the app is ready to run (exit 1 on a failure)
webcortex status          the running app's live view: requests, errors, runs, spend, devices
webcortex explain CODE    what an error code means; --list for all of them
webcortex trace [ID]      what a recent run did, as a span tree; `latest` for the newest
webcortex eval            run the declared evals; --record | --replay | --live
webcortex halt | release  the emergency stop for actuators

Security

WebCortex has been through an adversarial review of its own controls — auth, authorization, injection, traversal, SSRF, exhaustion, disclosure — plus stress and soak testing. Six issues were found and fixed, each with a regression test in tests/test_pentest.py:

Severity
Remote DoS + total auth failure via a JWT library panic Critical
No panic boundary on the request path High
Proxy path traversal usable as an SSRF primitive High
Unbounded behaviour recursion exhausting the worker pool High
Python tracebacks returned to clients Medium
Client input faults reported as 500s Low

Soak: 1,786,805 requests, 0 errors, 0 panics, memory at steady state.

SECURITY.md has the full report — including what was not tested, the known limits, and what v2 added to the surface.

Status

v2.9.1. Working and tested: the manifest IR, router, native ops (static / query / proxy / page / files / flow), the free-threaded Python bridge, authentication and scopes, rate limiting, CORS, security headers, graceful shutdown, Behaviours with concurrent leaves, the agent runtime with handoffs, sessions, resumable approval gates, composing budgets, tool-result bounding and compaction, context providers, memory, flows, two providers (Anthropic and OpenAI-compatible, which covers local models), prompt caching, the spend ledger, the audit trail, OpenAPI, the MCP server, TypeScript generation, the context pack and evolve, images in and out of agent runs and MCP, actuators and the emergency stop, the device hub (cameras, sensors, watchers, WebSocket ingest and streams, agents over WebSocket), diagnostics (stable error codes with hints on every failure, webcortex doctor, the live /_webcortex/status view and its MCP tool), and, new in 2.8, structured agent output validated by the runtime, provider request options, run traces with OTLP export, and evals that replay recorded model responses. 2.9 is a hardening release from a planned stress test of 2.8: 70 defects fixed or documented, three of them high severity (see the changelog and the review in SECURITY.md). 2.9.1 fixes what machine-driven testing of 2.9.0 found: about 1.2 million fuzzed requests, a differential test of the output validator against jsonschema, and property and mutation tests. 1,442 tests (369 Rust, including 28 property tests, and 1,073 Python, including a 54-test adversarial suite, property and mutation-killing tests, and offline end-to-end suites that drive the whole agent stack over HTTP), clippy clean.

Not yet: Postgres, token-level SSE streaming, durable agent runs that survive a restart. See DESIGN.md for the roadmap, honest risk grading, and — just as importantly — what is deliberately not being built.

For AI coding tools

AGENTS.md is the machine-facing reference: the complete API surface with exact signatures and defaults, the binding and scope rules, the constraints the runtime enforces, and the specific mistakes that are cheap to make and expensive to debug. Claude Code, Cursor, Codex, Aider and Copilot Workspace all read it by convention.

It is written to be correct rather than welcoming. Humans should start with the docs site instead.

Building from source

uv venv --python 3.13
uv pip install maturin pytest
.venv/bin/maturin develop --uv
.venv/bin/python -m pytest tests/

The free-threaded build is also supported and is the configuration the bridge was designed around:

uv venv --python 3.14t

License

Apache-2.0

Metadata

Release files for web-cortex-framework 2.9.1

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for web-cortex-framework 2.9.1
File Size Uploaded Provenance Actions
web_cortex_framework-2.9.1.tar.gz 479.9 kB With provenance Details for web_cortex_framework-2.9.1.tar.gz

Built distributions (wheels)

Table of built distributions (wheels) for web-cortex-framework 2.9.1
File Actions
web_cortex_framework-2.9.1-cp314-cp314t-win_amd64.whl CPython 3.14 CPython 3.14 free-threading Windows x86-64 Details for web_cortex_framework-2.9.1-cp314-cp314t-win_amd64.whl
web_cortex_framework-2.9.1-cp314-cp314t-manylinux2014_x86_64.manylinux_2_17_x86_64.whl CPython 3.14 CPython 3.14 free-threading Linux glibc 2.17+ x86-64 Details for web_cortex_framework-2.9.1-cp314-cp314t-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
web_cortex_framework-2.9.1-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.whl CPython 3.14 CPython 3.14 free-threading Linux glibc 2.17+ ARM64 Details for web_cortex_framework-2.9.1-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.whl
web_cortex_framework-2.9.1-cp314-cp314t-macosx_11_0_arm64.whl CPython 3.14 CPython 3.14 free-threading macOS 11.0+ ARM64 Details for web_cortex_framework-2.9.1-cp314-cp314t-macosx_11_0_arm64.whl
web_cortex_framework-2.9.1-cp314-cp314t-macosx_10_12_x86_64.whl CPython 3.14 CPython 3.14 free-threading macOS 10.12+ x86-64 Details for web_cortex_framework-2.9.1-cp314-cp314t-macosx_10_12_x86_64.whl
web_cortex_framework-2.9.1-cp314-cp314-win_amd64.whl CPython 3.14 CPython 3.14 Windows x86-64 Details for web_cortex_framework-2.9.1-cp314-cp314-win_amd64.whl
web_cortex_framework-2.9.1-cp314-cp314-manylinux2014_x86_64.manylinux_2_17_x86_64.whl CPython 3.14 CPython 3.14 Linux glibc 2.17+ x86-64 Details for web_cortex_framework-2.9.1-cp314-cp314-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
web_cortex_framework-2.9.1-cp314-cp314-manylinux2014_aarch64.manylinux_2_17_aarch64.whl CPython 3.14 CPython 3.14 Linux glibc 2.17+ ARM64 Details for web_cortex_framework-2.9.1-cp314-cp314-manylinux2014_aarch64.manylinux_2_17_aarch64.whl
web_cortex_framework-2.9.1-cp314-cp314-macosx_11_0_arm64.whl CPython 3.14 CPython 3.14 macOS 11.0+ ARM64 Details for web_cortex_framework-2.9.1-cp314-cp314-macosx_11_0_arm64.whl
web_cortex_framework-2.9.1-cp314-cp314-macosx_10_12_x86_64.whl CPython 3.14 CPython 3.14 macOS 10.12+ x86-64 Details for web_cortex_framework-2.9.1-cp314-cp314-macosx_10_12_x86_64.whl
web_cortex_framework-2.9.1-cp313-cp313-win_amd64.whl CPython 3.13 CPython 3.13 Windows x86-64 Details for web_cortex_framework-2.9.1-cp313-cp313-win_amd64.whl
web_cortex_framework-2.9.1-cp313-cp313-manylinux2014_x86_64.manylinux_2_17_x86_64.whl CPython 3.13 CPython 3.13 Linux glibc 2.17+ x86-64 Details for web_cortex_framework-2.9.1-cp313-cp313-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
web_cortex_framework-2.9.1-cp313-cp313-manylinux2014_aarch64.manylinux_2_17_aarch64.whl CPython 3.13 CPython 3.13 Linux glibc 2.17+ ARM64 Details for web_cortex_framework-2.9.1-cp313-cp313-manylinux2014_aarch64.manylinux_2_17_aarch64.whl
web_cortex_framework-2.9.1-cp313-cp313-macosx_11_0_arm64.whl CPython 3.13 CPython 3.13 macOS 11.0+ ARM64 Details for web_cortex_framework-2.9.1-cp313-cp313-macosx_11_0_arm64.whl
web_cortex_framework-2.9.1-cp313-cp313-macosx_10_12_x86_64.whl CPython 3.13 CPython 3.13 macOS 10.12+ x86-64 Details for web_cortex_framework-2.9.1-cp313-cp313-macosx_10_12_x86_64.whl
web_cortex_framework-2.9.1-cp312-cp312-win_amd64.whl CPython 3.12 CPython 3.12 Windows x86-64 Details for web_cortex_framework-2.9.1-cp312-cp312-win_amd64.whl
web_cortex_framework-2.9.1-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.whl CPython 3.12 CPython 3.12 Linux glibc 2.17+ x86-64 Details for web_cortex_framework-2.9.1-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
web_cortex_framework-2.9.1-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.whl CPython 3.12 CPython 3.12 Linux glibc 2.17+ ARM64 Details for web_cortex_framework-2.9.1-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.whl
web_cortex_framework-2.9.1-cp312-cp312-macosx_11_0_arm64.whl CPython 3.12 CPython 3.12 macOS 11.0+ ARM64 Details for web_cortex_framework-2.9.1-cp312-cp312-macosx_11_0_arm64.whl
web_cortex_framework-2.9.1-cp312-cp312-macosx_10_12_x86_64.whl CPython 3.12 CPython 3.12 macOS 10.12+ x86-64 Details for web_cortex_framework-2.9.1-cp312-cp312-macosx_10_12_x86_64.whl

Total release size: 130.3 MB

Release files / web_cortex_framework-2.9.1.tar.gz

Download URL web_cortex_framework-2.9.1.tar.gz
Size 479.9 kB
Tags Source
SHA-256 checksum
How to use checksums
540b59f4a0948e1383c6a2054864e56d320b7143e9f0cd4748e0c70dea5bda48
BLAKE2b-256 checksum
How to use checksums
db5f62460e2743b5f86d38a332bc8ac1c69a151a2c800740281ff644083c7f2a
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314t-win_amd64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314t-win_amd64.whl
Size 6.6 MB
Tags CPython 3.14 CPython 3.14 free-threading Windows x86-64
SHA-256 checksum
How to use checksums
9ca3bb5925bde88119d0196f08bf737104597bc68649c56d854cc075599a3990
BLAKE2b-256 checksum
How to use checksums
11042006717284834affa5051ba37d0bf6d8cbb1c416727858bd10ff9e880863
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314t-manylinux2014_x86_64.manylinux_2_17_x86_64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314t-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
Size 6.7 MB
Tags CPython 3.14 CPython 3.14 free-threading Linux glibc 2.17+ x86-64
SHA-256 checksum
How to use checksums
402ea2727f47d72c296f4e3b0130b5aeffe7281cf8f3985599fe10313c51f1e7
BLAKE2b-256 checksum
How to use checksums
ea8d8c6ba61d97eddb56f187d9da05f7eb6bd57713beb56da532fd6a6c660daf
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314t-manylinux2014_aarch64.manylinux_2_17_aarch64.whl
Size 6.5 MB
Tags CPython 3.14 CPython 3.14 free-threading Linux glibc 2.17+ ARM64
SHA-256 checksum
How to use checksums
6e77f3fac1f20da30a0a7b21a8c586154dd12f2a338ad1cf6a7852b3301a4b0c
BLAKE2b-256 checksum
How to use checksums
5353623f61859050521124a93093525fe1513a276a4ec636f1e943ab7c473d84
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314t-macosx_11_0_arm64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314t-macosx_11_0_arm64.whl
Size 6.1 MB
Tags CPython 3.14 CPython 3.14 free-threading macOS 11.0+ ARM64
SHA-256 checksum
How to use checksums
fac0995ebf5e6683dd3c17a764d1b719549c0d3ba4c346b9144014d68582a9e7
BLAKE2b-256 checksum
How to use checksums
9623b9745edbfef851b5ec8fda5a170806c381976263ee27a00e78d7a3b440f8
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314t-macosx_10_12_x86_64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314t-macosx_10_12_x86_64.whl
Size 6.5 MB
Tags CPython 3.14 CPython 3.14 free-threading macOS 10.12+ x86-64
SHA-256 checksum
How to use checksums
5c0273801e8cba89cd72d64daaf17f29f376c890e449c6896676dd2899e1a9e3
BLAKE2b-256 checksum
How to use checksums
023f4d837f225d69fe24e2bbf32930eb89b0c00002b478eb0ad24b2a8e6b8209
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314-win_amd64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314-win_amd64.whl
Size 6.6 MB
Tags CPython 3.14 Windows x86-64
SHA-256 checksum
How to use checksums
257719eb7c78783bd8f5cdb3928d192463a172e4553485d0ef664c9a51620dd6
BLAKE2b-256 checksum
How to use checksums
0f9e8d273abe2eecc559f3ca1c1a82b9c7b2d5df4caa9cd478633d0b50f10d7e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314-manylinux2014_x86_64.manylinux_2_17_x86_64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
Size 6.7 MB
Tags CPython 3.14 Linux glibc 2.17+ x86-64
SHA-256 checksum
How to use checksums
2a14b292f68578fc8d3a621b19d8ce55e881f9768f226f38cabe718afb1e15f0
BLAKE2b-256 checksum
How to use checksums
f1410c2197d3f704006d67a0bfe49a10fc73a577dbd8ad1785bee8985efe7912
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314-manylinux2014_aarch64.manylinux_2_17_aarch64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314-manylinux2014_aarch64.manylinux_2_17_aarch64.whl
Size 6.5 MB
Tags CPython 3.14 Linux glibc 2.17+ ARM64
SHA-256 checksum
How to use checksums
408f09ec4cb1fed4960917a9ec1816906c58817e861076898f4548a1c6e8f968
BLAKE2b-256 checksum
How to use checksums
21861941b568e3ac20a7c13efc71c81725aa6e0822474c0ca53a9c4a8321ea29
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314-macosx_11_0_arm64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314-macosx_11_0_arm64.whl
Size 6.1 MB
Tags CPython 3.14 macOS 11.0+ ARM64
SHA-256 checksum
How to use checksums
6082cc48e2dbfafe4e87f0b5002c44909f3678204a866f647b13b78bc40b8c2f
BLAKE2b-256 checksum
How to use checksums
8a1e3c994a8e0191ab982187d81024c9472157257d1925fa35197b1d25c615ad
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp314-cp314-macosx_10_12_x86_64.whl

Download URL web_cortex_framework-2.9.1-cp314-cp314-macosx_10_12_x86_64.whl
Size 6.5 MB
Tags CPython 3.14 macOS 10.12+ x86-64
SHA-256 checksum
How to use checksums
f28857441525e164664c862242c0a031d6cf7f011b619e2ee59a04a1b202d950
BLAKE2b-256 checksum
How to use checksums
3365362b51ea5d8ab1095171c9ddfdbce52fd7baf9113465be14c86bc845c83f
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp313-cp313-win_amd64.whl

Download URL web_cortex_framework-2.9.1-cp313-cp313-win_amd64.whl
Size 6.6 MB
Tags CPython 3.13 Windows x86-64
SHA-256 checksum
How to use checksums
2166fa3f859f83007031ec007c7535e6751922be6862c752873687aacbce30d1
BLAKE2b-256 checksum
How to use checksums
baa1275a6557f42a4a2b4442c4d9e52153a5f6a40af5ee3c44c64161d87b84f6
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp313-cp313-manylinux2014_x86_64.manylinux_2_17_x86_64.whl

Download URL web_cortex_framework-2.9.1-cp313-cp313-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
Size 6.7 MB
Tags CPython 3.13 Linux glibc 2.17+ x86-64
SHA-256 checksum
How to use checksums
eff61fd7e499e3d5385c74c169df10d7f825ad7ff7f227b24cbc6e082fc2a13c
BLAKE2b-256 checksum
How to use checksums
6c3796e32a9716232a4b5b0454cb5068b333f7d97f4bfdc029fcc1a988cfc518
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp313-cp313-manylinux2014_aarch64.manylinux_2_17_aarch64.whl

Download URL web_cortex_framework-2.9.1-cp313-cp313-manylinux2014_aarch64.manylinux_2_17_aarch64.whl
Size 6.5 MB
Tags CPython 3.13 Linux glibc 2.17+ ARM64
SHA-256 checksum
How to use checksums
a025a0359f64525e7f57b22e66d361ee99369a3867eab696ebf56ed80ac2ca18
BLAKE2b-256 checksum
How to use checksums
814d994ca6503f211d70624ec4924cfb70c8a13e6e96e02326f4bcf93c782a15
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp313-cp313-macosx_11_0_arm64.whl

Download URL web_cortex_framework-2.9.1-cp313-cp313-macosx_11_0_arm64.whl
Size 6.1 MB
Tags CPython 3.13 macOS 11.0+ ARM64
SHA-256 checksum
How to use checksums
32c41e122ad2727ffbf4ea882d08e8d3f608d52a711b1432ca1671d9e1e70599
BLAKE2b-256 checksum
How to use checksums
95b5ee05b4e4ac7e8e80a77869cba4d83ff0dd1bf745dfd134141d6d246a9f8e
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp313-cp313-macosx_10_12_x86_64.whl

Download URL web_cortex_framework-2.9.1-cp313-cp313-macosx_10_12_x86_64.whl
Size 6.5 MB
Tags CPython 3.13 macOS 10.12+ x86-64
SHA-256 checksum
How to use checksums
dcfbb86e158f1f126bfec68cd01e470cc66fb3d1ae7153e04a5f4cdfaa106bdd
BLAKE2b-256 checksum
How to use checksums
9c1cdc960fd31bd8ae1783dfb88083c47e11a2098520e2e4ca6e3fd49f0d055d
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp312-cp312-win_amd64.whl

Download URL web_cortex_framework-2.9.1-cp312-cp312-win_amd64.whl
Size 6.6 MB
Tags CPython 3.12 Windows x86-64
SHA-256 checksum
How to use checksums
8bf9736d647460cc77550c3f10ee4a593da6e7d8ec7818dec5481b8ecb18fb4f
BLAKE2b-256 checksum
How to use checksums
83df3849b9ff3f5c1eeb4c0d8bb910e800da927da23cf57aaca3b4f6400d1c65
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.whl

Download URL web_cortex_framework-2.9.1-cp312-cp312-manylinux2014_x86_64.manylinux_2_17_x86_64.whl
Size 6.7 MB
Tags CPython 3.12 Linux glibc 2.17+ x86-64
SHA-256 checksum
How to use checksums
4f04b9160c153453e28a77ede0c08e77d5c2067cf8485014b0c7d9fbde63aed4
BLAKE2b-256 checksum
How to use checksums
95032d00e49e3e1b5d2239e0dfb78bd9211306d99a49466a62e5291d1124e293
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.whl

Download URL web_cortex_framework-2.9.1-cp312-cp312-manylinux2014_aarch64.manylinux_2_17_aarch64.whl
Size 6.5 MB
Tags CPython 3.12 Linux glibc 2.17+ ARM64
SHA-256 checksum
How to use checksums
11b76c4277dcee3c499727b55fe96a4ebe01738f7096e508e317d79ec4296615
BLAKE2b-256 checksum
How to use checksums
c2d589b6b845a4ee6cd290bd5530c866c0023885662d2dc5dd7503a93c75b2f0
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp312-cp312-macosx_11_0_arm64.whl

Download URL web_cortex_framework-2.9.1-cp312-cp312-macosx_11_0_arm64.whl
Size 6.1 MB
Tags CPython 3.12 macOS 11.0+ ARM64
SHA-256 checksum
How to use checksums
90f6bc255a95a91c53d4e7a741740407ebecdb5fa84fed591fe4ef90d0ff6824
BLAKE2b-256 checksum
How to use checksums
723368f4340f1e9706d809f69574382375efa90219d1257801b0c5b6a339a7ff
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Release files / web_cortex_framework-2.9.1-cp312-cp312-macosx_10_12_x86_64.whl

Download URL web_cortex_framework-2.9.1-cp312-cp312-macosx_10_12_x86_64.whl
Size 6.5 MB
Tags CPython 3.12 macOS 10.12+ x86-64
SHA-256 checksum
How to use checksums
d9ba347b3b4fb73e1e7cbf14606c4cf7bd519681c45148a7772f7a32cfae82d1
BLAKE2b-256 checksum
How to use checksums
63789f2f844aba00dd2d7379f9d7ff220d58f04b7d6e18e01dfda055b517fb13
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

Security for web-cortex-framework 2.9.1

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

All 21 files in this release have been uploaded with provenance metadata.

Select from 21 files

Attestation for web_cortex_framework-2.9.1.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 22, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page