Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/eric-cielo/moflo/memory-worktreenpx skills add eric-cielo/moflo --skill memory-worktreegit clone --depth 1 https://github.com/eric-cielo/mofloWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00134 | $0.01467 |
| Opus 5 | $0.00067 | $0.00733 |
| Sonnet 5 | $0.00027 | $0.00293 |
| Haiku 4.5 | $0.00013 | $0.00147 |
Grade A, and why
memory-worktree scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 128 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/memory-worktree — Durable Sharing Across Worktrees (Automatic)
Working across git worktrees is a core moflo competency, so durable-learning sharing across the
worktrees of a repo is now automatic and on by default. When a checkout is part of a repo with
worktrees in play, moflo derives a shared durable store at <git-common-dir>/moflo/durable.db and
converges the learnings + knowledge namespaces across every worktree at session start — no config.
A plain single checkout writes nothing and behaves exactly as before.
This skill is for the three things you might still want: verify it's working, customize where it stores (or extend it to non-worktree clones), or opt out.
Only the durable slice travels. Structural + ephemeral data stays local per checkout and rebuilds itself — the same safety model the auto path is built on (a dedicated
durable.db, never a fullmoflo.db). See [[memory-team]] for the git-tracked, cross-machine variant.
Arguments: $ARGUMENTS — --off to opt out, --path <store> to set a custom shared store.
When to use
- "Is memory / are learnings shared across my worktrees?" → verify (Step 1).
- "Where are my worktree learnings stored?" → it's
<git-common-dir>/moflo/durable.db; verify to confirm. - "Turn off / disable worktree memory sharing" → opt out (Step 2).
- "Share learnings across my separate clones / a custom location" → customize (Step 3).
For a team, or your own laptop+desktop over git, that's /memory-team, not this.
Procedure
Step 0 — Memory first (gate requirement)
mcp__moflo__memory_search { query: "durable_path worktree sharing auto git common dir learnings", namespace: "guidance" }
Step 1 — Verify (default action)
Run the safety/status check and surface it verbatim:
npx flo doctor -c shared-db 2>&1
- "Automatic worktree learning sharing active ()" → it's working; the path is the shared store.
- "No shared full moflo.db detected…" → auto sharing is NOT active here. Most likely this checkout has no sibling worktrees yet (auto only turns on once worktrees exist). Confirm with the user; if they expected sharing across separate clones (not worktrees), go to Step 3 — auto can't converge those.
- A
warn→ a full-DB misconfig is present; follow the check'sfixline.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 128 lines · 134 tokens per session scan A 07f0d689a9f8
memory-worktree is a skill published in the GitHub repository eric-cielo/moflo (18 stars, last pushed 4d ago), licensed MIT. It adds 134 tokens to every session and 1,467 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
fleet-manager
Use when managing, triaging, restarting, escalating, or summarizing Codewhale Pod runs and workers.
gh-assign-issues
Use to assign GitHub issues to a milestone and/or owners in bulk, verifying each.
mcp-builder
Design, build, configure, or debug Model Context Protocol servers for codewhale, including stdio and HTTP/SSE transports.
feishu
Work with Feishu or Lark bots, docs, sheets, bitables, approval flows, and OpenAPI/MCP setup without hardcoding credentials.
interview
Ask one useful structured question at a time only when material product/implementation choices are genuinely missing; remember answers and produce a brief/spec. Discoverable facts should be investigated instead of asked.
security-review
Review trust boundaries, auth/authz, injection, secrets, filesystem/network exposure, dependencies, and exploitability without pretending a shallow lint is an audit.