Commands

Every command supports --json for structured output. Global flags: -v/--version, -q/--quiet, --verbose, --timing. ANSI colors respect NO_COLOR.

| Command | Description | |---------|-------------| | ml init | Initialize .mulch/ in the current project | | ml add <domain> | Add a new expertise domain | | ml record <domain> --type <type> | Record an expertise record (--tags, --force, --relates-to, --supersedes, --batch, --stdin, --dry-run, --evidence-bead) | | ml edit <domain> <id> | Edit an existing record by ID or 1-based index | | ml delete <domain> [id] | Delete records by ID, --records <ids>, or --all-except <ids> (--dry-run) | | ml delete-domain <domain> | Remove a domain from config and delete its expertise JSONL file (--yes, --dry-run) | | ml query [domain] | Query expertise (--all, --classification, --file, --outcome-status, --sort-by-score, --format filters) | | ml prime [domains...] | Output AI-optimized expertise context (--budget, --no-limit, --context, --files, --exclude-domain, --format, --export) | | ml search [query] | Search records across domains with BM25 ranking (--domain, --type, --tag, --classification, --file, --sort-by-score, --format) | | ml compact [domain] | Analyze compaction candidates or apply a compaction (--analyze, --auto, --apply, --dry-run, --min-group, --max-records) | | ml diff [ref] | Show expertise changes between git refs (ml diff HEAD~3, ml diff main..feature) | | ml status | Show expertise freshness and counts (--json for health metrics) | | ml validate | Schema validation across all files | | ml doctor | Run health checks on expertise records (--fix to auto-fix) | | ml setup [provider] | Install provider-specific hooks (claude, cursor, codex, gemini, windsurf, aider) | | ml onboard | Generate AGENTS.md/CLAUDE.md snippet | | ml prune | Remove stale tactical/observational entries | | ml ready | Show recently added or updated records (--since, --domain, --limit) | | ml sync | Validate, stage, and commit .mulch/ changes | | ml outcome <domain> <id> | Append an outcome to a record (--status, --duration, --agent, --notes), or view outcomes | | ml upgrade | Upgrade mulch to the latest version (--check for dry run) | | ml learn | Show changed files and suggest domains for recording learnings | | ml completions <shell> | Output shell completion script (bash, zsh, fish) |

Architecture

Mulch stores expertise as typed JSONL records in .mulch/expertise/<domain>.jsonl — one file per domain, one record per line. Six record types (convention, pattern, failure, decision, reference, guide) with three classification tiers (foundational, tactical, observational) govern shelf life and pruning. Advisory file locks and atomic writes ensure safe concurrent access from multiple agents. Schema validation (via Ajv) enforces type-specific required fields. See CLAUDE.md for full technical details.

How It Works

1. ml init               → Creates .mulch/ with domain JSONL files
2. Agent reads expertise     → Grounded in everything the project has learned
3. Agent does work           → Normal task execution
4. Agent records insights    → Before finishing, writes learnings back to .mulch/
5. git push                  → Teammates' agents get smarter too

The critical insight: step 4 is agent-driven. Before completing a task, the agent reviews its work for insights worth preserving and calls ml record. Mulch provides the schema and file structure so those learnings land in a consistent, queryable format.

What's in .mulch/

.mulch/
├── expertise/
│   ├── database.jsonl        # All database knowledge
│   ├── api.jsonl             # One JSONL file per domain
│   └── testing.jsonl         # Each line is a typed, structured record
└── mulch.config.yaml         # Config: domains, governance settings

Everything is git-tracked. Clone a repo and your agents immediately have the project's accumulated expertise.

Record Types

| Type | Required Fields | Use Case | |------|----------------|----------| | convention | content | "Use WAL mode for SQLite connections" | | pattern | name, description | Named patterns with optional file references | | failure | description, resolution | What went wrong and how to avoid it | | decision | title, rationale | Architectural decisions and their reasoning | | reference | name, description | Key files, endpoints, or resources worth remembering | | guide | name, description | Step-by-step procedures for recurring tasks |

All records support optional --classification (foundational / tactical / observational), evidence flags (--evidence-commit, --evidence-issue, --evidence-file, plus tracker-specific --evidence-bead, --evidence-seeds, --evidence-gh, --evidence-linear), --tags, --relates-to, --supersedes for linking, and --outcome-status (success/failure) for tracking application results. Cross-domain references use domain:mx-hash format (e.g., --relates-to api:mx-abc123). When evidence.commit or files[] are omitted, ml record auto-populates them from the current git context.

Example Output

$ ml query database

## database (6 records, updated 2h ago)

### Conventions
- Use WAL mode for all SQLite connections
- Migrations are sequential, never concurrent

### Known Failures
- VACUUM inside a transaction causes silent corruption
  → Always run VACUUM outside transaction boundaries

### Decisions
- **SQLite over PostgreSQL**: Local-only product, no network dependency acceptable

Design Principles

• Zero LLM dependency — Mulch makes no LLM calls. Quality equals agent quality.
• Provider-agnostic — Any agent with bash access can call the CLI.
• Git-native — Everything lives in .mulch/, tracked in version control.
• Append-only JSONL — Zero merge conflicts, trivial schema validation.
• Storage != Delivery — JSONL on disk, optimized markdown/XML for agents.

Concurrency & Multi-Agent Safety

Mulch is designed for multi-agent workflows where several agents record expertise concurrently against the same repository.

How it works

• Advisory file locking — Write commands acquire a .lock file (O_CREAT|O_EXCL) before modifying any JSONL file. Retries every 50ms for up to 5 seconds; stale locks (>30s) are auto-removed.
• Atomic writes — All JSONL mutations write to a temp file first, then atomically rename into place. A crash mid-write never corrupts the expertise file.
• Git merge strategy — ml init sets merge=union in .gitattributes so parallel branches append-merge JSONL lines without conflicts.

Command safety

| Safety level | Commands | Notes | |---|---|---| | Fully safe (read-only) | prime, query, search, status, validate, learn, ready | No file writes. Any number of agents, any time. | | Safe (locked writes) | record, edit, delete, delete-domain, compact, prune, doctor | Acquire per-file lock before writing. Multiple agents can target the same domain — the lock serializes access automatically. | | Serialize (setup ops) | init, add, onboard, setup | Modify config or external files (CLAUDE.md, git hooks). Run once during project setup, not during parallel agent work. |

Swarm patterns

Same-worktree agents (e.g., Claude Code team, parallel CI jobs):

# Every agent can safely do this in parallel:
ml prime                                    # Read context
ml record api --type pattern --name "..." --description "..."  # Locked write
ml search "error handling"                  # Read-only

Locks ensure correctness automatically. If two agents record to the same domain at the same instant, one waits (up to 5s) for the other to finish.

Multi-worktree / branch-per-agent:

Each agent works in its own git worktree. On merge, merge=union combines all JSONL lines. Run ml doctor --fix after merge to