synix 0.13.0

Build system for agent memory

 ███████╗██╗   ██╗███╗   ██╗██╗██╗  ██╗
 ██╔════╝╚██╗ ██╔╝████╗  ██║██║╚██╗██╔╝
 ███████╗ ╚████╔╝ ██╔██╗ ██║██║ ╚███╔╝
 ╚════██║  ╚██╔╝  ██║╚██╗██║██║ ██╔██╗
 ███████║   ██║   ██║ ╚████║██║██╔╝ ██╗
 ╚══════╝   ╚═╝   ╚═╝  ╚═══╝╚═╝╚═╝  ╚═╝

A build system for agent memory.

The Problem

Agent memory hasn't converged. Mem0, Letta, Zep, LangMem — each bakes in a different architecture because the right one depends on your domain and changes as your agent evolves. Most systems force you to commit to a schema early. Changing your approach means migrations or starting over.

What Synix Does

Conversations are sources. Prompts are build rules. Summaries and world models are artifacts. Declare your memory architecture in Python, build it, then change it — only affected layers rebuild. Trace any artifact back through the dependency graph to its source conversation.

uvx synix build pipeline.py
uvx synix search "return policy"
uvx synix validate                # experimental

Quick Start

uvx synix init my-project
cd my-project

Add your API key (see pipeline.py for provider config), then build:

uvx synix build

Browse, search, and validate:

uvx synix list                    # all artifacts, grouped by layer
uvx synix show final-report       # render an artifact
uvx synix search "hiking"         # full-text search
uvx synix validate                # run declared validators (experimental)

Defining a Pipeline

A pipeline is a Python file. Layers are real objects with dependencies expressed as object references.

# pipeline.py
from synix import Pipeline, Source, SearchIndex
from synix.ext import MapSynthesis, ReduceSynthesis

pipeline = Pipeline("my-pipeline")
pipeline.source_dir = "./sources"
pipeline.build_dir = "./build"
pipeline.llm_config = {
    "provider": "anthropic",
    "model": "claude-haiku-4-5-20251001",
    "temperature": 0.3,
    "max_tokens": 1024,
}

# Parse source files
bios = Source("bios", dir="./sources/bios")

# 1:1 — apply a prompt to each input
work_styles = MapSynthesis(
    "work_styles",
    depends_on=[bios],
    prompt="Infer this person's work style in 2-3 sentences:\n\n{artifact}",
    artifact_type="work_style",
)

# N:1 — combine all inputs into one output
report = ReduceSynthesis(
    "report",
    depends_on=[work_styles],
    prompt="Write a team analysis from these profiles:\n\n{artifacts}",
    label="team-report",
    artifact_type="report",
)

pipeline.add(bios, work_styles, report)
pipeline.add(SearchIndex("search", sources=[work_styles, report], search=["fulltext"]))

This is a complete, working pipeline. uvx synix build pipeline.py runs it.

For the full pipeline API, built-in transforms, validators, and advanced patterns, see docs/pipeline-api.md.

Configurable Transforms (synix.ext)

Most LLM steps follow one of four patterns. The synix.ext module provides configurable transforms for each — no custom classes needed.

from synix.ext import MapSynthesis, GroupSynthesis, ReduceSynthesis, FoldSynthesis

Transform	Pattern	Use when...
MapSynthesis	1:1	Each input gets its own LLM call
GroupSynthesis	N:M	Group inputs by a metadata key, one output per group
ReduceSynthesis	N:1	All inputs become a single output
FoldSynthesis	N:1 sequential	Accumulate through inputs one at a time

All four take a prompt string with placeholders like {artifact}, {artifacts}, {group_key}, {accumulated}. Changing the prompt automatically invalidates the cache.

For full parameter reference and examples of each, see docs/pipeline-api.md#configurable-transforms.

When you need logic beyond prompt templating — filtering, conditional branching, multi-step chains — write a custom Transform subclass.

Built-in Transforms

Pre-built transforms for common agent memory patterns. Import from synix.transforms:

Class	What it does
EpisodeSummary	1 transcript → 1 episode summary
MonthlyRollup	Group episodes by month, synthesize each
TopicalRollup	Group episodes by user-defined topics
CoreSynthesis	All rollups → single core memory document
Merge	Group artifacts by content similarity (Jaccard)

CLI Reference

Command	What it does
uvx synix init <name>	Scaffold a new project with sources, pipeline, and README
uvx synix build	Run the pipeline. Only rebuilds what changed
uvx synix plan	Dry-run — show what would build without running transforms
uvx synix plan --explain-cache	Plan with inline cache decision reasons
uvx synix list [layer]	List all artifacts, optionally filtered by layer
uvx synix show <id>	Display an artifact. Resolves by label or ID prefix. --raw for JSON
uvx synix search <query>	Full-text search. --mode hybrid for semantic
uvx synix validate	(Experimental) Run validators against build artifacts
uvx synix fix	(Experimental) LLM-assisted repair of violations
uvx synix lineage <id>	Show the full provenance chain for an artifact
uvx synix clean	Delete the build directory
uvx synix batch-build plan	(Experimental) Dry-run showing which layers would batch vs sync
uvx synix batch-build run	(Experimental) Submit a batch build via OpenAI Batch API. --poll to wait
uvx synix batch-build resume <id>	(Experimental) Resume a previously submitted batch build
uvx synix batch-build list	(Experimental) Show all batch build instances and their status
uvx synix batch-build status <id>	(Experimental) Detailed status for a specific batch build. --latest for most recent

Batch Build (Experimental)

Warning: Batch build is experimental. Commands, state formats, and behavior may change in future releases.

The OpenAI Batch API processes LLM requests asynchronously at 50% cost with a 24-hour SLA. Synix wraps this into batch-build — submit your pipeline, disconnect, come back when it's done.

Quick Example

# pipeline.py — mixed-provider pipeline
pipeline.llm_config = {
    "provider": "openai",           # OpenAI layers → batch mode (automatic)
    "model": "gpt-4o",
}

episodes = EpisodeSummary("episodes", depends_on=[transcripts])
monthly = MonthlyRollup("monthly", depends_on=[episodes])

# Force this layer to run synchronously via Anthropic
core = CoreSynthesis("core", depends_on=[monthly], batch=False)
core.config = {"llm_config": {"provider": "anthropic", "model": "claude-sonnet-4-20250514"}}

# Submit and wait for completion
uvx synix batch-build run pipeline.py --poll

Poll vs Resume

Poll workflow — submit and wait in a single session:

uvx synix batch-build run pipeline.py --poll --poll-interval 120

Resume workflow — submit, disconnect, come back later:

# Submit (exits after first batch is submitted)
uvx synix batch-build run pipeline.py
#   Build ID: batch-a1b2c3d4
#   Resume with: synix batch-build resume batch-a1b2c3d4 pipeline.py --poll

# Check on it later
uvx synix batch-build status --latest

# Resume and poll to completion
uvx synix batch-build resume batch-a1b2c3d4 pipeline.py --poll

The batch Parameter

Each transform accepts an optional batch parameter controlling whether it uses the Batch API:

Value	Behavior
None (default)	Auto-detect: batch if the layer's provider is native OpenAI, sync otherwise.
True	Force batch mode. Raises an error if the provider is not native OpenAI.
False	Force synchronous execution, even if the provider supports batch.

episodes = EpisodeSummary("episodes", depends_on=[transcripts])              # auto
monthly = MonthlyRollup("monthly", depends_on=[episodes], batch=True)        # force batch
core = CoreSynthesis("core", depends_on=[monthly], batch=False)              # force sync

Provider Restrictions

Batch mode only works with native OpenAI (provider="openai" with no custom base_url). Transforms using Anthropic, DeepSeek, or OpenAI-compatible endpoints via base_url always run synchronously. Setting batch=True on a non-OpenAI layer is a hard error.

Transform Requirements

Transforms used in batch builds must be stateless — their execute() method must be idempotent and produce deterministic prompts from the same inputs. All built-in transforms (EpisodeSummary, MonthlyRollup, TopicalRollup, CoreSynthesis) meet this requirement.

See docs/batch-build.md for the full specification including state management, error handling, and the request collection protocol.

Key Capabilities

Incremental rebuilds — Change a prompt or add new sources. Only downstream artifacts reprocess.

Full provenance — Every artifact chains back to the source conversations that produced it. uvx synix lineage <id> shows the full tree.

Fingerprint-based caching — Build fingerprints capture inputs, prompts, model config, and transform source code. Change any component and only affected artifacts rebuild. See docs/cache-semantics.md.

Altitude-aware search — Query across episode summaries, rollups, or core memory. Drill into provenance from any result.

Architecture evolution — Swap monthly rollups for topic-based clustering. Transcripts and episodes stay cached. No migration scripts.

Where Synix Fits

	Mem0	Letta	Zep	LangMem	Synix
Approach	API-first memory store	Agent-managed memory	Temporal knowledge graph	Taxonomy-driven memory	Build system with pipelines
Incremental rebuilds	—	—	—	—	Yes
Provenance tracking	—	—	—	—	Full chain to source
Architecture changes	Migration	Migration	Migration	Migration	Rebuild
Schema	Fixed	Fixed	Fixed	Fixed	You define it

Synix is not a memory store. It's the build system that produces one.

Learn More

Doc	Contents
Pipeline API	Full Python API — ext transforms, built-in transforms, projections, validators, custom transforms
Entity Model	Artifact identity, storage format, cache logic
Cache Semantics	Rebuild trigger matrix, fingerprint scheme
Batch Build	(Experimental) OpenAI Batch API for 50% cost reduction
CLI UX	Output formatting, color scheme

Links

• synix.dev
• GitHub
• llms.txt — machine-readable project summary for LLMs
• Issue tracker — known limitations and roadmap
• MIT License