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.
git clone --depth 1 https://github.com/fusengine/agentsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/fusengine/agents/lessons)<a href="https://agentmods.dev/commands/fusengine/agents/lessons"><img src="https://agentmods.dev/badge/commands/fusengine/agents/lessons/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/fusengine/agents/lessons"><img src="https://agentmods.dev/badge/commands/fusengine/agents/lessons.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00026 | $0.00550 |
| Opus 5.5 | $0.00010 | $0.00220 |
| Sonnet 5.5 | $0.00005 | $0.00110 |
| Haiku 4.5 | $0.00003 | $0.00055 |
Grade A, and why
lessons 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 1mo 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 — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/lessons
Manage the project's lessons memory — a MEMORY/LESSON.md file committed at the project root. It records compact post-mortems so the same mistake is never reproduced twice.
Architecture
MEMORY/LESSON.md— the committed list of lessons. One lesson per line, newest at the bottom. This is the file you read and append to.MEMORY/state.json— the throttle counter (last-reminder / last-write timestamps). It is gitignored and managed by the hooks; never edit it by hand.
The fuse-lessons hooks force-read MEMORY/LESSON.md at every session and sub-agent start, so anything written here is automatically reloaded into context.
View existing lessons
Read MEMORY/LESSON.md at the project root (create it if absent). Each line is one lesson, newest at the bottom.
Add a lesson
Append one line in this exact compact format:
- [YYYY-MM-DD HH:MM] <what went wrong> → <what to do instead>
Rules:
- One lesson per line. Keep it short — a single actionable sentence.
- Timestamp uses the current local date and time in
YYYY-MM-DD HH:MM. TheStophook injects the exact stamp to use — copy it verbatim. - Left of
→: the concrete failure (what broke, what was wrong). - Right of
→: the corrective action (what to do next time). - No paragraphs, no code blocks — just the bullet line.
Example
- [2026-06-20 11:02] Added "hooks": "./hooks/hooks.json" string in marketplace entry → never set a hooks field in marketplace.json; hooks auto-load by convention from hooks/hooks.json
Throttle
The Stop hook reminds you to capture a lesson at most once every 5 minutes by default. Override with the FUSE_LESSONS_THROTTLE_MIN environment variable (minutes). Writing to MEMORY/LESSON.md resets the throttle automatically via the PostToolUse hook (it updates MEMORY/state.json).
After appending, mention that the lesson will be force-read on the next session/sub-agent start, and that MEMORY/LESSON.md should be committed with the change so the whole team inherits it.
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.
- 1mo ago First seen · 47 lines · 26 tokens per session scan A 557d62397709
lessons is a command published in the GitHub repository fusengine/agents (33 stars, last pushed 9d ago), licensed MIT. It adds 26 tokens to every session and 550 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-09-03.
Other commands, from other repositories
brain
Set up or use your Professional Brain — init, ingest, recall, or review.
pin
Break checkpoint — status display, session log, optional MEMORY.md update.
on
Turn on senior mode — refuses vague questions, hints instead of answers, never writes code.
unwrap
You are resuming a Claude Code session. Execute all steps below in order — no confirmation needed.
continue
Resume an interrupted autoimprove session from where it left off.
checklist
Generate a custom checklist for the current feature based on user requirements.