agent-cerebro 0.4.1

Persistent two-tier memory for AI agents. Short-term markdown + long-term SQLite with semantic search.

Agent Cerebro

Persistent two-tier memory for AI agents. Battle-tested across 134 sessions with 10 agent roles.

Short-term (markdown files, always loaded) + Long-term (SQLite + OpenAI embeddings, searched on-demand).

Install

pip install agent-cerebro

Zero required dependencies. SQLite is Python stdlib.

Optional semantic search:

pip install agent-cerebro[embeddings]
export OPENAI_API_KEY="sk-..."

Quick Start

CLI

# Initialize
cerebro init

# Store a memory (auto-dedup via cosine similarity >0.92)
cerebro store coder gotchas "kamal app exec spawns new container, use docker exec"
cerebro store social exhausted_stories "blue-green deploy order loss" --tags deploy,sqlite

# Search (semantic + keyword fallback)
cerebro search coder gotchas "kamal file not found"
cerebro search coder gotchas "deploy issue" --tag critical

# List categories
cerebro list coder

# Timeline — chronological view of all memories
cerebro timeline coder
cerebro timeline coder --last 7d
cerebro timeline coder --last 2w --category gotchas

# Export — dump all memories for a role
cerebro export coder --format md > coder_memories.md
cerebro export coder --format json > coder_memories.json
cerebro export coder --format json --category gotchas

# Stats — storage metrics and category breakdown
cerebro stats
cerebro stats coder

# Garbage collection — find and remove near-duplicates
cerebro gc coder --dry-run
cerebro gc coder --apply
cerebro gc coder --threshold 0.85 --category gotchas

# Check health
cerebro check --all

Python API

from agentrecall import MemoryStore, MemorySearch, MemoryTimeline, MemoryExport, MemoryStats, MemoryGC

# Store
store = MemoryStore()
store.store("coder", "gotchas", "kamal spawns new container", tags=["kamal", "docker"])

# Search (with optional tag filter)
search = MemorySearch()
results = search.search("coder", "gotchas", "kamal file not found")
results = search.search("coder", "gotchas", "deploy issue", tag="critical")

# Timeline
timeline = MemoryTimeline()
entries = timeline.timeline("coder", last="7d")

# Export
export = MemoryExport()
markdown = export.export("coder", fmt="md")
json_str = export.export("coder", fmt="json", category="gotchas")

# Stats
stats = MemoryStats()
metrics = stats.stats(role="coder")
# → {total_entries, total_with_embeddings, embedding_coverage_pct, db_size_bytes, ...}

# Garbage collection
gc = MemoryGC()
result = gc.gc("coder", dry_run=True)
# → {found: 3, removed: 0, duplicates: [...]}
result = gc.gc("coder", dry_run=False)  # actually delete

How It Works

Two-Tier Design

Short-term (memory/<role>.md)	Long-term (SQLite + embeddings)
Active learnings, mistakes, feedback	Growing lists (exhausted topics, defect patterns)
Max 80 lines, pruned regularly	Unlimited entries, never pruned
Read in full at session start	Searched on-demand per query

Semantic Dedup

Every store call embeds the text via OpenAI text-embedding-3-small and checks cosine similarity against all existing entries in the same role/category. Similarity > 0.92 blocks the store (raises DuplicateError).

Without an API key, falls back to exact text matching.

Search

1. Embed the query
2. Compute cosine similarity against all entries with embeddings
3. Return entries above threshold (0.75), sorted by similarity
4. If no embedding matches: keyword fallback (>=50% keyword match)
5. No API key (or embeddings unavailable): keyword-only search
6. Optional --tag filter narrows results to entries with a specific tag
7. Optional --limit N caps output to the top-N results (keeps agent context tight)

Garbage Collection

cerebro gc finds near-duplicate entries within each role/category pair:

• With embeddings: cosine similarity >= threshold (default 0.92)
• Without embeddings: exact text match (case-insensitive)
• Older entry (lower ID) is kept; newer duplicate is removed
• --dry-run (default) reports without deleting
• --apply actually removes duplicates

Graceful Degradation

Works fully offline without an OpenAI API key:

• Store: exact text dedup (case-insensitive)
• Search: keyword matching (>=50% of query words must appear)
• GC: exact text match dedup only

Network Resilience

Embedding calls are wrapped in a retry loop with exponential backoff + jitter. Transient failures (timeouts, dropped connections, HTTP 429/5xx) are retried; HTTP 429 honors the Retry-After header. After exhausting retries — or on a non-retryable error (e.g. 401) — store/search print a warning and fall back to keyword/exact-match instead of crashing mid-session. Auth/4xx errors are not retried (no point burning attempts on a bad key).

Tune via CEREBRO_EMBED_MAX_RETRIES (default 3) and CEREBRO_EMBED_TIMEOUT (seconds, default 30).

Agent Skills

Copy skill/agent-recall/ into your project's skills directory for use with Claude Code, Codex, Cursor, Copilot, Cline, or Goose.

cp -r skill/agent-recall/ .claude/skills/agent-recall/

Configuration

Environment variables:

Variable	Default	Description
AGENT_CEREBRO_HOME	~/.agent-cerebro	Memory storage directory
OPENAI_API_KEY	(none)	OpenAI API key for embeddings
UT_OPENAI_API_KEY	(none)	Preferred over OPENAI_API_KEY
CEREBRO_EMBED_MAX_RETRIES	3	Retries on transient embedding-API failures
CEREBRO_EMBED_TIMEOUT	30	Per-request embedding timeout (seconds)

CLI Reference

cerebro store <role> <category> "text" [--tags t1,t2] [--db path]
cerebro search <role> <category> "query" [--tag tagname] [--limit N] [--db path]
cerebro list <role> [--db path]
cerebro timeline <role> [--last 7d] [--category cat] [--limit N] [--db path]
cerebro export <role> [--format md|json] [--category cat] [--db path]
cerebro stats [role] [--db path]
cerebro gc <role> [--dry-run] [--apply] [--threshold 0.92] [--category cat] [--db path]
cerebro check [--fix] [--long-term] [--all] [--dir path] [--db path]
cerebro init [--dir path]
cerebro migrate [--dry-run] [--rebuild] [--dir path] [--db path]

agentrecall and agentmemory also work as CLI aliases.

Exit codes: 0 = success/found, 1 = not-found/validation-fail, 2 = input error.

Migration from JSONL

If you have existing JSONL memory files:

cerebro migrate --dir /path/to/memory/
cerebro migrate --rebuild  # Re-embed entries missing embeddings

Related Tools

Part of the Ultrathink Agent Suite:

• Agent Architect Kit — Multi-agent starter kit that uses Cerebro for cross-session memory
• Agent Orchestra — Task queue + orchestration CLI for spawning and managing agents
• AgentBrush — Image editing toolkit for AI agents

Built by an AI-run dev shop. Read how →

Newsletter

stdout — a free newsletter on running AI agents in production: memory, orchestration, failure modes, real P&L. From the team that dogfoods Cerebro daily.

License

MIT