agentleFS
leninejuniorleninejunior/engrene-memory-bridge/AGENTS.md

engrene-memory-bridge

AGENTS.md12 starsChanged 56 days agoMITView on GitHub

Welcome, AI Agent! This project uses Memory Bridge (engrene-memory-bridge) alongside Compound Engineering (CE) principles to maintain seamless context, project standards, and historical learnings across all AI tools. Whenever operating in this repository, follow this execution lifecycle:

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 31,542 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository4 other files it gives its agents.

The file

We split this AGENTS.md into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before making any changes, inspect active context:

    line 20
  2. Distinctive:

    When making non-trivial architectural choices or adding dependencies:

    line 26
  3. Distinctive:

    Before wrapping up your turn:

    line 32
  4. Distinctive:

    Keep memory portable between tools

    line 49
  5. Distinctive:

    Add stable context that should persist across sessions.

    line 52

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

AGENTS.md — Universal AI Agent Guidelines (Compound Engineering)

Welcome, AI Agent! This project uses Memory Bridge (engrene-memory-bridge) alongside Compound Engineering (CE) principles to maintain seamless context, project standards, and historical learnings across all AI tools.


🧭 Memory Bridge Lifecycle Protocol (Mandatory)

Whenever operating in this repository, follow this execution lifecycle:

flowchart LR
    A["1. Resume Context\n(memory-bridge resume)"] --> B["2. Search / Inspect\n(memory-bridge search)"]
    B --> C["3. Execute & Record Decisions\n(memory-bridge decision add)"]
    C --> D["4. Build & Verify\n(npm test)"]
    D --> E["5. Log & Handoff\n(memory-bridge log && handoff build)"]

1. Session Start

Before making any changes, inspect active context:

memory-bridge resume --for <agent-name>

2. Decision Recording

When making non-trivial architectural choices or adding dependencies:

memory-bridge decision add --title "Title" --decision "Choice" --context "Rationale" --impact "Effects"

3. Verification & Handoff

Before wrapping up your turn:

npm run build && npm test
memory-bridge log --tool <agent-name> --intent "Summary" --summary "Details"
memory-bridge handoff build

📌 Project Invariants & Context

Project Context

Current Objective

Describe the current milestone and expected outcome.

Constraints

  • Local-first only
  • Keep memory portable between tools

Notes

Add stable context that should persist across sessions.


🏛️ Active Architectural Decisions & Project Pitfalls

  • [dec-mu7tvkhc] Independent Community Memory Bridge: Adopt filesystem JSONL and Markdown contracts under .memory-bridge with zero runtime dependencies (Impact: Full transparency, no vendor lock-in, cross-tool continuity between Claude, Codex, Gemini, Antigravity and Orca)
  • [dec-mu7tvkjh] Hybrid Local Search Engine: Combine BM25 term weighting with local embedded SQLite cosine similarity (Impact: Fast, local, offline search without cloud LLM API dependencies)
  • [dec-muc3rsh5] Obsidian Vault Export & JEV BM25 Hybrid Provider Support: Added optional Obsidian vault sync (memory-bridge obsidian --vault <path>) exporting Markdown notes with YAML frontmatter and wikilinks, and added 'jev' semantic provider option for joint code-text search combined with BM25 SQLite FTS5. (Impact: Expands memory-bridge export capabilities to Obsidian and improves semantic vector retrieval options.)
  • [dec-mucoine1] Compound Engineering (CE) Multi-Agent Rules Synchronization: Added 'memory-bridge ce' CLI command and core module (src/core/ce.ts) to automatically generate and sync AGENTS.md, .cursorrules, CLAUDE.md, and copilot-instructions.md with active decisions and pitfalls. (Impact: Ensures all AI assistants automatically inherit project decisions, memory lifecycle protocols, and pitfalls without manual prompt editing.)
  • [dec-mucplvwt] Automatic Multi-Agent & Obsidian Sync Triggers: Added automatic integration execution (src/core/auto-sync.ts) triggered on 'decision add', 'log', and 'handoff build' events to keep Compound Engineering files and Obsidian vaults in sync automatically. (Impact: Eliminates manual sync steps and keeps rule files and Obsidian notes updated in real-time.)
  • [dec-mudge6ri] Obsidian vault import is bidirectional, vault wins by default: Added 'memory-bridge obsidian import' and 'obsidian sync' (import then export). Decisions match by frontmatter id; notes without id get a generated id written back and export reuses the user's file name. Sessions match by ts+tool. Conflicts resolved by --prefer vault (default) or bridge; updates replace the JSONL record in place (same id, no duplicate). Redaction runs on import. config.integrations.obsidian.autoImport pulls before resume. loadConfig now preserves the integrations block (it was silently dropped before, so vaultDir/autoSync from config.json never applied). (Impact: src/core/obsidian.ts (parser + importObsidianVault), src/core/store.ts (replaceDecisionEvent/replaceSessionEvent), src/core/config.ts (normalizeIntegrations), CLI obsidian subcommands, SPEC §8.1. Handoff.md stays export-only.)
  • [dec-mudi62tf] This repository commits its own .memory-bridge/ (selective .gitignore): Track config.json, decisions.jsonl, sessions/*.jsonl, handoff.md and project-context.md in Git; ignore only vector.sqlite, .lock and observations/. Every log/decision/handoff change is committed together with the code. redaction.customPatterns redacts IPv4(:port); hostnames and SSH aliases are not covered, so never put server names in summaries (the repo is public). (Impact: .gitignore, src/core/config.ts ensureGitignoreEntry, src/core/redaction.ts detectLeakageRisk, README portability section. Working tree gets dirty after any memory write; commit it.)

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.