remaind 0.9.14

Local-first memory and context continuity for AI agents working beyond one context window.

Remaind

Local-first continuity and handoff infrastructure for AI work across sessions, models, agents, and tools.

Remaind is a local-first context ledger, compaction pipeline, resume substrate, and universal handoff system for AI agents, local models, OpenAI-compatible API models, coding agents, and automation workflows. It lets agents handle work that is larger than a model's live context window by preserving the meaningful state of the run across resets and handoffs: a future agent can start with a clean model context, load Remaind's local state, understand what happened before, know what must happen next, and continue safely without asking the user to reconstruct the work.

Remaind is not a wiki — it is a machine-readable and human-readable continuity and handoff layer: raw event ledger, structured state, compact handover, searchable memory, structured validation, safe rollback, a mechanical resume gate, and model-neutral handoff bundles that another LLM or harness can read.

Remaind also treats retrieved memory as evidence, not instructions. The resume gate uses mechanical, defense-in-depth heuristics to flag known hostile instruction shapes before a fresh model sees retrieved memories or raw excerpts, and the compaction validator rejects poisoned local model output before it can become durable memory. These checks improve hygiene around untrusted content; they are not a standalone security boundary.

New events are also hash-chained in their metadata. remaind validate verifies that chain and reports tampering when a chained event edit, deletion, or reorder breaks continuity. Resume packets label retrieved events and memories with mechanical trust hints such as user, verified_tool, assistant, or source_linked_memory.

Install

python3 -m pip install --upgrade remaind
remaind --version
remaind launch

Requires Python ≥ 3.11. The install gives you the remaind terminal command, local memory/state storage, context packets, plan reconciliation, rollback, monitoring, large-document tools, and model-agnostic launch/run flows.

Remaind does not install Python, local model runtimes, model files, or API credentials. For model-generated answers, choose one of these at launch:

• a local Ollama model
• a local OpenAI-compatible server such as llama.cpp, vLLM, or LM Studio
• an API-backed OpenAI-compatible endpoint, including Qwen-style services
• packet-only / no-model workflows when you only need memory, context inspection, or recovery artifacts

If python3 is missing, install Python 3.11+ first, then run the install command above.

Sovereign Memory Starter Kit

Remaind now ships a portable product kit for users who want a complete local agent memory setup rather than only the Python library:

starter-kits/sovereign-memory-starter/

The starter kit adds:

• one-command bootstrap/setup, readiness checks, and launch
• managed Python bootstrap through local, checksum-verified uv when Python 3.11+ is missing
• model profiles for Qwen, DeepSeek, DeepSeek v4 Flash, GLM, Minimax-style, Llama, and custom local models
• Ollama and OpenAI-compatible local compaction/chat, including llama.cpp server
• Qwen Code and built-in console adapters
• a synthetic continuity proof (make proof) that verifies fresh-process retrieval
• a beyond-context benchmark (make bench) for distant contradiction, revocation pressure, source-link, and adversarial checks
• a direct large-prompt runner (remaind run FILE "TASK") that builds a compact evidence packet before calling the configured local model
• a model battle-test harness (make model-test) for deterministic Remaind retrieval checks plus optional local model matrix testing
• privacy-safe sharing: each user gets their own .context/ ledger

Quick start from the repo:

cd starter-kits/sovereign-memory-starter
./bootstrap.sh --profile qwen-large --agent qwen-code
remaind

Console-only local proof:

cd starter-kits/sovereign-memory-starter
./bootstrap.sh --agent console --profile llama
make proof

Build the distributable zip:

scripts/build-sovereign-memory-starter.sh

This writes a release archive under dist/.

See the full product guide in docs/sovereign-memory-starter.md. For the day-one workflow of starting a product and later continuing from memory without manual packet management, see docs/zero-to-continuation.md.

Quick start

# Install or upgrade from PyPI.
python3 -m pip install --upgrade remaind

# Check the installed version.
remaind --version

# Run the 5-minute deterministic proof path.
remaind proof

# Optional: prove a real local DeepSeek can continue from a Remaind packet.
export REMAIND_OPENAI_BASE_URL=http://127.0.0.1:11442/v1
export REMAIND_OPENAI_MODEL=deepseek-v4-flash
export REMAIND_OPENAI_API_KEY="$(cat /path/to/local-deepseek-api-key)"
export REMAIND_OPENAI_API=completions
remaind demo agent-continuity --force

# Start the guided workflow after the proof passes.
remaind launch

# Or bootstrap a .context/ in the current directory manually.
remaind init

# Inspect what's there.
remaind validate
remaind status
remaind doctor

# One front-door startup: product/workspace lookup first, model/runtime
# selection next, then ask, paste, file, resume, handover, or empty mode.
remaind launch

# After work happens (events appended by your agent harness),
# compact when token band climbs.
remaind compact

# Build a resume packet for a fresh agent run.
remaind resume --next-tool deploy_prod

# Start a guided Codex -> DeepSeek transfer with capture + receiver guide.
remaind transfer start \
  --from-model "Codex" \
  --to-model "local DeepSeek" \
  --target codex \
  --completed "wired retry route" \
  --decision "use provider + event_id for idempotency" \
  --next-step "run focused retry tests"

# After the receiving model writes its final output, capture it back.
remaind transfer finish \
  --from-model "local DeepSeek" \
  --from-file deepseek-output.md

# Register a huge local Markdown/text file, then retrieve exact evidence
# by search or line range without loading the whole file into the prompt.
remaind large-doc ingest /path/to/large-archive.md
remaind large-doc search /path/to/large-archive.md "final synthesis phrase" --verify
remaind large-doc slice /path/to/large-archive.md --start 33557 --end 33565 --verify

# Or execute a task against a huge local prompt/document in one command.
remaind run /path/to/huge-prompt.md "solve this task"

# Roll back if something went wrong.
remaind rollback --to 2026-05-14T03:54:33Z

Use it as a library

import remaind is the stable public API — remaind.__all__ is the whole surface. The underscore-prefixed modules are internal.

import remaind

base = "./my-agent-run"
remaind.init(base)

state = remaind.status(base)
writer = remaind.EventWriter.open(base)
writer.append(remaind.EventInput(
    type="user_message", actor="user", summary="Asked to refactor auth",
    session_id=state["session_id"], task_id=state["task_id"],
    content="Please refactor src/auth/...", importance=3,
))

if remaind.compaction_status(base).compaction_needed:
    remaind.compact(base)   # uses your local model if one is running

packet = remaind.resume(base).packet   # inject packet.content into a fresh context

Full walkthrough — the agent-loop integration, how compaction uses your local model, the resume-injection pattern, and an exception reference — is in docs/integration.md.

Adversarial memory behavior is documented in docs/adversarial-hardening.md. Executable product proofs are documented in docs/benchmarks.md. The fastest public trust path is remaind proof: it creates isolated local proof workspaces, verifies the Basic/offline plan boundary, runs the project-continuity benchmark, runs the sovereign beyond-context benchmark, verifies universal capture/handoff export, and prints one PASS/FAIL report with the claims each check proves. remaind proof --secure adds the secure workspace lock/unlock, encrypted core-storage, and encrypted-handoff proof. The proof contract is documented in docs/proofs/golden-proof-path.md. The finance policy-exception continuity walkthrough is in docs/proofs/finance-policy-exception-continuity.md.

Large Local Documents

Agent terminals can read files that agent UI file-readers often refuse to load. Remaind exposes that as a product feature through remaind large-doc. For the guided terminal workflow, remaind launch activates memory, verifies runtime readiness, and only then asks whether to ask a project question, paste text, load a file, prepare a resume packet, or start empty:

remaind launch
remaind launch --project "Rutwais"
remaind launch --path ./my-project --input ask --task "status of Rutwais?"
remaind context inspect --project "Rutwais" "what is the current plan and next step?"
remaind launch --path ./my-project --input file --file /absolute/path/to/huge.md
cat /absolute/path/to/huge.md | remaind launch --path ./my-project --input paste --task "solve this task"

In an interactive terminal, remaind launch opens a session home first. Use the Up/Down keys and Enter to choose New session, Open by name or topic, Continue last project, one of the last 10 recent projects, the current folder, or a manual path. New sessions get their own timestamped workspace folder by default, so each product session can keep a dedicated .context/.

When you choose Open by name or topic, type something like Rutwais or local llm test. Remaind builds a workspace-memory index from existing .context folders, then searches folder names, paths, handover, state, recent packets, and recent events. The lookup tolerates hyphens, close spellings, and related topic phrases such as merchant refunds. Only strong matches open automatically; weak or ambiguous matches are shown as selectable full paths. If nothing matches, launch offers to create a new session named after what you typed, browse existing projects, or try another name. Remaind remembers safe project-name aliases, but ignores generic shell phrases such as which remaind so Recent Projects stay readable.

Activate memory and return to terminal and Prepare resume packet are memory-only flows and do not probe, load, or call a model. Model/runtime selection is delayed until you choose a mode that needs generation, such as asking Remaind memory, loading a file, pasting text, or updating handover. The empty terminal path reports Runtime: not selected for this mode; that means model selection was skipped for speed, not that DeepSeek or another configured model is unavailable. The model picker lists detected local Ollama/OpenAI-compatible models before the readiness check. If a selected model fails its smoke test, launch lets you choose another detected model before offering packet-only fallback. File mode lists registered large documents and nearby Markdown/text/JSON-style files. Use the visible Back row to return to the previous step without restarting. Press m for a manual path when the path you need is not listed. The continuation picker also includes Update handover.md; in terminal selectors, Ctrl+H activates that handover update shortcut.

For the closest thing to always-on memory, install the daemon shell hook:

eval "$(remaind daemon shell-init --shell zsh)"

Or install it once:

remaind daemon install-shell-hook --shell zsh

The hook keeps the active-memory prompt hint and records privacy-safe work signals after each command: sanitized command preview, exit code, rough duration, current folder, git branch, and changed files. It never records terminal output. Meaningful developer commands and failed developer commands update the product status and product-plan reconciliation automatically, so Remaind can recover the latest activity and map it back to the active plan even if the terminal closes unexpectedly. On zsh, the hook also routes plain-English questions typed at an active-memory prompt through remaind ask before zsh treats them as shell commands. For example, after opening a workspace, is the repo cloned locally? is converted to a Remaind memory question instead of producing a shell glob error. The last workspace opened through remaind launch becomes the active memory target for these follow-up questions.

After choosing a workspace, the default path is normal terminal/LLM usage: press Enter and choose Activate memory and return to terminal to activate memory and return control without putting Remaind between you and the model. You are back in your shell after this path; plain natural-language questions should be asked through your normal LLM interface or with remaind ask "your question". Type memory-specific questions such as what is the next step?, latest progress and TODOs, or ask what did we finish last? when you want Remaind to build context and answer from memory. Type prepare memory packet, update handover, paste, file, or a document path to route to those explicit Remaind workflows. Generic LLM prompts typed at this point are saved as lightweight launch notes and returned to the normal session instead of being processed before the LLM.

On MOSAI/DeepSeek installs that include the profile switcher, launch also shows a DeepSeek mode picker before model selection. Choose Turbo - faster, 65k context for day-to-day speed or Stable - larger, 131k context when the task needs the widest local context. The switch restarts the supervised DeepSeek server, updates Remaind's packet/context settings, and then continues through the normal readiness checks.

If you do not have a local LLM, choose Connect API model (OpenAI-compatible) from the model picker. Remaind asks for the service's /v1 endpoint, accepts an optional bearer token for the current launch only, lists the API models it can discover, and then runs the same readiness check before any project question or large prompt is processed. The API key is not written into .context/.

Ask Remaind memory mode runs your question against the current resume packet. It is intentionally a memory-QA path, not the default chat path. If you choose Prepare resume packet only, interactive launch offers to ask a question from that packet before returning to the shell, so natural text like status of Rutwais? is captured by Remaind instead of being interpreted as a terminal command. Follow-up questions in the same launch carry the current terminal conversation forward, so corrections like “the model is already working” are treated as newer than stale packet failures. If you ask for recent packets, launch includes the latest run packets and answer notes as evidence for that question. More generally, memory questions now include a workspace packet inventory so the model knows which prior packets and answer notes exist. Normal questions include recent packet context; requests for all packets or full history load every packet and answer note that fits the fixed question-source budget and clearly list anything omitted. For speed, launch answers operational checks such as “is Remaind active?” and terminal-command inputs such as gh repo clone ... locally instead of sending them through the model. Model-backed memory questions use a launch-specific timeout cap, REMAIND_LAUNCH_ASK_MODEL_TIMEOUT (default 45 seconds), so a slow or stalled local model cannot leave the interactive chat waiting for the general model timeout. The same context-engineering trace is now part of all model-facing context: resume packets, remaind run retrieval-planner prompts and evidence packets, handover compaction prompts, and deep-document compaction prompts all label selected evidence by authority, freshness, and budget status before a model sees it. When your goal is to resume work, use normal language such as continue working on this project, what is the next step?, or latest progress and TODOs. Remaind treats that as a continuation request: the LLM receives the resume packet plus budgeted packet-library context and is asked to reconstruct the product scope, active plan, latest progress, blockers, next TODOs, and the immediate implementation step before it answers. On activation, Remaind also seeds a durable product-status record when one does not exist yet. That seed is intentionally conservative: it says memory is active, no product progress or blocker has been captured yet, and the next answer should use the workspace plan, handover, and packet evidence to choose the real next step. Remaind also writes .context/active/plan_reconciliation.md and .context/active/plan_reconciliation.json, which turn loaded plan/status files plus saved activity into a plain-English map of completed, blocked, in-progress, pending, and planned work. Continuation answers use that reconciliation before old packets or old model answers.

For advanced local setups, each detected model is available in normal mode and YOLO mode, for example deepseek-v4-flash and deepseek-v4-flash (YOLO / skips model test). That lets you choose the model without typing or thinking about provider endpoints. The Model not listed entry is only for unusual cases where a local model exists but does not appear in discovery. The rest of readiness still runs, and the launch report labels the model as unverified. The same path is available in scripts with:

remaind launch --runtime ollama --model any-local-model:latest --yolo-model

Paste mode is intentionally available after readiness. It captures pasted or stdin text into .context/artifacts/launch-pastes/ and routes that captured file through the same large-prompt packet path, so a long terminal session can start naturally without pushing raw text into the model before memory is active.

For the shortest one-shot workflow, remaind run FILE "TASK" combines initialization, ingest, retrieval, packet writing, local-model execution, and event logging:

remaind run /absolute/path/to/huge-prompt.md "solve this task"
remaind run /absolute/path/to/huge-prompt.md "solve this task" --deep-compact

The source file stays on disk. Remaind writes a compact evidence packet to .context/active/run_packet.md, writes the model response to .context/active/run_answer.md, and records the task in the append-only event log.

For million-token class files, add --deep-compact. Remaind first checkpoints a baseline packet, then uses the selected local model to extract source-linked chunk summaries, validates quote and line bounds, quarantines hostile derived content, reduces the evidence, and expands exact raw lines for answer-affecting claims. If a local model request stalls or is interrupted, the baseline packet and pending answer file already exist, so the run does not leave you empty handed. Lenient runs with unrecovered chunks report partial; strict mode fails on any unrecovered chunk.

Slower local runtimes can use smaller extraction chunks for faster checkpoint cadence:

remaind run /absolute/path/to/huge-prompt.md "solve this task" --deep-compact --deep-compact-chunk-tokens 2000

The chunk target is part of the deep-compaction cache key. Reuse with --reuse-deep-compact only picks up an interrupted or completed run with the same source, task, model, prompts, settings, and chunk target. The default target is 8000 tokens; lower it for timeout-prone local runtimes and raise it only for stable long-context models.

Deep-compaction artifacts live under .context/large-docs/<doc_id>/deep-compactions/<run_id>/ and can be inspected or pruned without copying the raw source into .context/:

remaind large-doc inspect /absolute/path/to/huge-prompt.md
remaind large-doc prune-deep-compactions --doc /absolute/path/to/huge-prompt.md --keep-latest 2

Use it when a source file is too large for a direct @file read, too large for one prompt, or too important to summarize without exact evidence:

remaind large-doc ingest /absolute/path/to/archive.md
remaind large-doc search /absolute/path/to/archive.md "needle phrase" --verify --json
remaind large-doc slice /absolute/path/to/archive.md --start 1000 --end 1040 --verify --json

ingest writes a durable local manifest under .context/large-docs/, appends a source-linked event, and stores a searchable memory pointer with the exact commands future agents should use. The source document is not copied into .context/; it remains at its original local path. search, slice, head, and tail stream the file through the terminal and return line-numbered evidence.

Use --verify when retrieving from an ingested document. Remaind compares the current file against the stored manifest and stops if the source has changed. If you launch an agent UI and type raw @/path/to/huge.md, that UI may still hit its own file-reader size limit; the Remaind path is the large-doc command, not direct @file loading.

This does not give a model one-shot full attention over every token. It gives the agent a reliable local way to retrieve the relevant slices from documents bigger than its live context or file-reader limit, then reason over those slices with the normal Remaind resume and validation loop.

What lives in .context/

.context/
├── README.md
├── CONTRACT.md           # the contract — read this first
├── active/
│   ├── state.json        # derived working state (atomic replace)
│   ├── handover.md       # compact continuity document (atomic replace)
│   └── (resume_packet.md, history/  — runtime, git-ignored)
├── logs/
│   └── events.jsonl      # append-only raw timeline (source of truth)
├── large-docs/           # runtime manifests for huge local documents
├── schemas/
│   ├── event.schema.json         # JSON Schema Draft 2020-12
│   ├── state.schema.json
│   ├── memory.schema.json
│   ├── validation.schema.json
│   ├── thresholds.yaml           # 40k/60k/70k/80k band math
│   ├── redaction.yaml            # 13 default secret patterns
│   ├── tools.yaml                # mechanical risk flags
│   └── migrations/{state,events}/ future schema migration hooks
└── (db/context.sqlite, artifacts/  — runtime, git-ignored)

Authority order

When sources disagree, lower wins:

1. Latest explicit user instruction
2. Raw event log (logs/events.jsonl)
3. active/state.json
4. active/handover.md
5. Derived memories

A stale summary or memory MUST NOT override a newer user instruction.

Commands

Command	What it does
remaind init	Bootstrap .context/; --force backs up existing
remaind init --secure	Bootstrap and seal a secure workspace: sensitive core files are encrypted at rest and plaintext handoff export is disabled
remaind unlock --ttl 30m	Decrypt sealed core files and open a bounded secure-handoff window
remaind lock	Reseal sensitive core files and close the secure-handoff window
remaind key rotate	Re-encrypt locked secure workspace storage with a new passphrase
remaind validate	Walk the v1 checklist (structure, schemas, event hash chain, SQLite)
remaind status [--json]	State + thresholds + event counts + compaction recommendation
remaind plan [--json]	Show the local Basic/Pro entitlement boundary; Phase 1 has no network, billing, or license checks
remaind doctor [--json]	Read-only health check: validation, status readability, and local model runtime detection
remaind monitor [--json]	Scheduler-friendly health and alert check; exits non-zero on invalid context, urgent compaction, or unhealthy deep-compaction runs
remaind launch [--project NAME] [--input ask|paste|file|resume|handover|empty]	Primary startup flow: open session home, create or activate project memory, delay runtime checks until generation is needed, then route into ask/paste/file/resume/handover/empty mode
remaind ask [--project NAME] "QUESTION"	Ask the active workspace memory in plain language
remaind chat [--project NAME]	Stay inside an interactive Remaind memory chat
remaind context inspect [--project NAME] "QUESTION"	Print the engineered memory context Remaind would send for a question, without calling a model
remaind context checkpoint --active-tokens N --context-window M	Model-agnostic context guard: create a fresh handoff/reset prompt when any agent reaches the 70% window guard
remaind compact [--no-model]	Run the compaction pipeline, gated by structured validation. Uses a local model automatically (Ollama auto-detected; OpenAI-compatible servers via REMAIND_OPENAI_BASE_URL), else the rule-based fallback. Use --no-model to force the deterministic fallback when a local runtime is slow or unhealthy
remaind resume [--next-tool TOOL]	Build a resume packet; consult the resume gate
remaind capture --from-file OUTPUT.md	Capture completed model/agent work into events, state, progress, and structured memory
remaind handoff --to-model MODEL --target codex	Build a universal cross-model handoff bundle with markdown packet, manifest, and adapter files
remaind transfer start --from-model Codex --to-model "local DeepSeek"	Build a guided cross-LLM transfer with receiver instructions, universal handoff bundle, manifest, adapters, and finish command
remaind transfer finish --from-model "local DeepSeek" --from-file OUTPUT.md	Capture the receiving model's completed work back into the same local continuity memory
remaind context checkpoint --agent qwen-code --model qwen --active-tokens 92000 --context-window 131072	Create the same tiny fast-start handoff for any agent/model with token telemetry; DeepCode is optional
remaind context checkpoint --telemetry-text "status: processing · tokens: 115904" --context-window 131072 --agent gemma-terminal --model gemma-4	Parse a plain terminal status line and create the same fast task-slice checkpoint without a custom adapter
remaind context watch --telemetry-file agent-status.log --context-window 131072 --agent gemma-terminal	Watch a generic agent status/log file and trigger the fast checkpoint automatically when the 70% guard is crossed
remaind context start	Print the smallest safe startup prompt for the latest handoff so the next agent reads the task slice first
remaind hooks status --agent deepcode --require-auto	Search local hook surfaces and verify that an external agent has a real automatic checkpoint hook; exits non-zero instead of claiming unsupported automation
remaind hooks install --agent deepcode --context-window 65536	Install and verify the native DeepCode checkpoint hook
remaind hooks probe --agent codex --require-discovered	Safely prove a discovered external hook contract can generate a checkpoint handoff using synthetic token telemetry
remaind hermes activate --workspace PATH	Install Hermes plugins and safely configure Hermes to use Remaind, with a config backup
remaind hermes activate --workspace PATH --verify --verify-report-dir reports	Activate Hermes, run installed-readiness, low-noise UX, and local proof-suite verification, then save Markdown/JSON evidence
remaind hermes deactivate	Remove Remaind's Hermes config selections with a backup, leaving unrelated settings intact
remaind hermes restore --backup PATH	Restore Hermes config from a Remaind-created backup, preserving the current config first
remaind hermes status --require-ready	Check Hermes plugin files, config, and workspace readiness before opening Hermes Desktop
remaind hermes speed --checkpoint-smoke --require-ready	Measure Hermes startup context size/time and write a real fast checkpoint/reset smoke proof
remaind hermes ux-proof --require-ready	Verify the recommended low-noise Hermes/Remaind launch path, prompt-size warnings, and checkpoint/reset proof
remaind hermes verify --force --report-dir reports	Verify installed Hermes readiness, low-noise UX readiness, and the local Remaind/Hermes proof suite, then save Markdown/JSON evidence
remaind hermes support-bundle --out reports/hermes --force --zip	Write a redacted Hermes/Remaind support bundle with status, UX proof, proof reports, config evidence, checksums, and an optional zip archive
remaind hermes support-bundle-check --bundle reports/hermes.zip	Verify a support bundle archive, sidecar hash, manifest checksums, UX proof evidence, and privacy flags before trusting it
remaind hermes install	Install or refresh Remaind's Hermes memory-provider and context-engine plugins under $HERMES_HOME
remaind deepcode install-hook --context-window 65536	Install a project-local DeepCode notify hook that uses the universal context guard against DeepCode's local token counter
remaind deepcode checkpoint --context-window 65536	Inspect DeepCode's local session index and prepare a reset prompt/handoff when the latest session is near the model window
remaind handoff --secure --ttl 30m	Encrypt the exported handoff bundle, write a secure manifest, and remove plaintext export files from the handoff folder
remaind handoff-decrypt --bundle secure_handoff.remaind.enc --out DIR	Decrypt a secure handoff bundle into an explicit folder for the receiving model or harness
remaind rollback --to <ts>	Restore derived files from history; raw log untouched
remaind memory plan|block|decision|promote|changed ...	Record plan updates, blockers, decisions, durable memories, and changes through the append-only memory layer
remaind run FILE "TASK"	Execute a task against a huge local prompt/document through a compact evidence packet and the selected local or OpenAI-compatible model runtime
remaind run FILE "TASK" --deep-compact	Run hierarchical local-model document compaction before answering; artifacts are resumable and source-linked
remaind large-doc ...	Index, search, and slice huge local documents through terminal-backed access without copying the source file into .context/
remaind demo finance-handoff	Run a product-facing finance handoff demo through a real local .context/ ledger and write a private customer validation artifact pack
remaind demo agent-continuity	Prove a loopback local DeepSeek model can continue a software-agent task from a Remaind resume packet and cite source event evidence
remaind proof	Run the golden clean-install proof path: plan boundary + project continuity + sovereign beyond-context + universal handoff checks
remaind proof --secure	Add secure workspace checks: plaintext export rejection, unlock, encrypted storage, encrypted handoff, decrypt, cleanup, key rotation, and lock
remaind bench sovereign	Run the beyond-context product proof in an isolated benchmark workspace
remaind bench context-engine --force	Prove strict fast-context budgets, active brief quality, stale-state boundaries, and forced-reset checkpoint behavior
remaind bench hermes-suite --force	Run the local Hermes proof battery: lifecycle, acceptance, quality, comparison, and long-run reset loop
remaind bench hermes-live	Verify the installed Hermes config, plugin files, provider tools, context engine, workspace resolution, and reset telemetry
remaind bench hermes-compare --force	Compare Remaind against no-memory/raw-transcript/built-in baselines and inventory installed Hermes providers
remaind bench hermes-quality --force	Prove Hermes receives current source-linked Remaind task context under stale/noisy long-history pressure
remaind bench hermes-lifecycle --force	Prove Hermes activate, deactivate, restore, dry-run, backup, and managed-file cleanup behavior
remaind bench hermes-longrun --force	Prove repeated Hermes/Remaind reset cycles beyond a 500k-token job load
remaind bench finance-handoff	Run the Oz-aligned finance policy-exception continuity proof
remaind bench finance-handoff --write-fixture-template customer.json	Write a customer-validation fixture template for private design-partner workflow labels
remaind bench finance-handoff --fixture customer.json	Run the deterministic finance proof against customer-provided reviewed-memory cases and gold labels

Plain telemetry accepts status text such as tokens: 115.9k/131k, tokens: 115904 of 131072, or status: processing · tokens: 115904. If the status line reports only the current token count, pass --context-window or set REMAIND_CONTEXT_WINDOW so Remaind can evaluate the 70% guard honestly.

Plan Boundary

Remaind ships as one local-first CLI. In Phase 1, remaind plan reports the offline entitlement boundary:

• Basic is the default and works offline forever.
• Basic includes local .context, local memory, resume packets, handoff read/export, basic continuity proof, secure local storage basics, and readable/exportable user data.
• Pro gates are prepared locally for advanced workflows, governed/reviewed memory, finance/compliance proof packs, richer reports, model integration automation, and future sync/backups/team features.
• No project content, memory, packets, briefs, handoffs, workspace names, or source notes are sent to any server in Phase 1.

Doctrine: Pro unlocks advanced workflows, not ownership of user memory.

The repo includes a runnable design-partner starter fixture at docs/fixtures/finance-policy-exception-design-partner-template.json. It is a schema and evaluator starter, not a real customer validation result.

The Sovereign Memory starter also includes make adversarial, a deterministic proof for prompt-injection, fake tool-result, phase-2 poison archives, superseded-memory, and poisoned compaction-output handling.

It also includes make bench, which generates a large source archive and proves fresh-process final-phrase retrieval, a single distant contradiction, revocation pressure, source links, and adversarial quarantine.

For a terminal-visible local model A/B demo, the starter kit includes make proof-2m: direct raw 2M prompt versus compact sovereign-memory resume packet against the configured local model.

For model compatibility work, the starter kit includes make model-test. Run it without model generation for a fast deterministic Remaind check:

make model-test MODEL_TEST_ARGS="--skip-model --force"

Run it against explicit local models:

make model-test MODEL_TEST_ARGS="--models qwen3.6-35b-a3b-full:latest,gpt-oss-120b-fast:latest --force"

Architecture

Phase	Subject
1	Frozen contract — schemas, configs, layout
2	Migration interfaces — state migrations + event adapters (Protocols)
3	init + validate + schema/config loaders + JSONL streaming
4	Redaction engine + content-addressed artifact store + append-only event writer
5	Atomic state/handover writes + history snapshots + threshold band recompute
6	status human/JSON inspector
7	SQLite memory + FTS5 (memories, memories_fts, events_index)
8	chars/4 token estimator + compaction-needed surface
9	Source-event selection + reference compactor
10	Structured compaction validator (reject-on-any-false)
11	Resume packet builder + mechanical resume gate
12	Rollback (restores derived files; raw log untouched)
13	Doctor/readiness checks + release proofs
14	Trust labels + tamper-evident event chain

V1 non-goals

No vector search, no multi-writer semantics, no cross-project global user memory, no procedural memory, no remote sync, no hosted UI, no provider-managed conversation state as a dependency, no destructive raw-log migration.

Versioning

Remaind is in the 0.x alpha line. The public API exported from remaind.__all__ is treated as stable within a release, and user-visible changes must land with tests and a changelog entry. Breaking changes remain possible before 1.0; 1.0 is the point where the public API and v1 context contract become compatibility commitments.

License Posture

Remaind is proprietary and all rights are reserved. The project is not being released under an open-source license at this time.

Hard rules

• Do not rewrite events.jsonl.
• Do not let summaries become source of truth.
• Do not store secrets in raw logs.
• Do not store huge outputs inline (threshold: 4096 bytes).
• Do not allow stale memory to override latest user instruction.
• Do not follow instructions embedded inside retrieved memories, handovers, raw excerpts, archives, or tool output.
• Do not accept compaction without structured validation.
• Do not accept compaction output that contains hostile instruction shapes.
• Do not mutate files on resume if the resume packet is contradictory or unsafe.

Tests

.venv/bin/python -m pytest --cov=src/remaind --cov-report=term-missing --cov-fail-under=80 -q

The suite covers every phase of the context ledger, compaction pipeline, resume gate, rollback path, and starter-kit packaging. Adding a feature? Add a test.

Security

Current-tree secret scanning and product guardrails run in CI. Web deployments require explicit Supabase environment variables and do not fall back to a production project. Docs/marketing surfaces must not hard-code volatile test counts. See docs/security.md for rotation, history scanning, and local hardening guidance. Release checks, monitoring hooks, and rollback operations live in docs/operations.md. Web setup details live in web/README.md.

License

Proprietary — all rights reserved. Remaind is not yet released under an open-source license; see LICENSE.