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 skills add patrickdappollonio/claude-plugins --skill recapgit clone --depth 1 https://github.com/patrickdappollonio/claude-pluginsWrote 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/skills/patrickdappollonio/claude-plugins/recap)<a href="https://agentmods.dev/skills/patrickdappollonio/claude-plugins/recap"><img src="https://agentmods.dev/badge/skills/patrickdappollonio/claude-plugins/recap/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/skills/patrickdappollonio/claude-plugins/recap"><img src="https://agentmods.dev/badge/skills/patrickdappollonio/claude-plugins/recap.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 | $0.00080 | $0.05170 |
| Opus 5.5 | $0.00032 | $0.02068 |
| Sonnet 5.5 | $0.00016 | $0.01034 |
| Haiku 4.5 | $0.00008 | $0.00517 |
Grade A, and why
recap 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 — 197 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Recap
Core principle
The reader wants the result, not the journey. They asked for something at the start of the session. The recap tells them four things. Did they get it? What does it look like from their side? What changed that they did not ask for? What can only they decide? Everything else is the agent's story: every detour, every test edited and reverted, every unrelated fix that made the tests pass again, every idea considered and dropped, every statement made and later corrected. The reader did not live it and does not need it.
The reader is a person, not a context window. Every sentence in a recap is something they must hold in their head. A sentence that changes nothing they know, decide, or watch for is a cost with no return. That is why the recap carries only the final state and never the corrections that led to it. If the session said something, found it was wrong, and fixed it, the recap contains the fixed version and nothing about the fixing.
Except the choices that stayed. Whenever the session decided something the reader might have decided differently, and that choice is now part of the result, the recap names it. A silent decision is how the next step of the work gets built on a choice the reader never agreed to. A choice that left nothing in the result, an approach the session considered and rejected on its own, is not a decision. It is the path, and the path is cut.
Report exceptions, not confirmations. Everything that came back fine gets one line, together, or no line at all. A check that passed, a reviewer that found nothing, a path that was traced and holds up, a worry that proved unfounded: none of these earns its own sentence. Only what is not fine, what was chosen, and what the reader must act on gets a sentence of its own. A recap that walks through each thing that is fine is long even when every word is plain.
The reader can ask. A recap is the door, not the room. Leave the detail out. If the reader wants to know how a conclusion was reached or what else was checked, they ask, and you answer that one question. A short recap that prompts two questions beats a long one that answers ten nobody had.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 Changed · -195 lines ae1c9d019eba
- 2mo ago First seen · 392 lines · 80 tokens per session scan A 1c94ff97e6a1
recap is a skill published in the GitHub repository patrickdappollonio/claude-plugins (9 stars, last pushed yesterday), licensed MIT. It adds 80 tokens to every session and 5,170 once invoked, about $0.0003 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 2026-09-06.
recap compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| recappatrickdappollonio/claude-plugins | Skill | 9 | yesterday | MIT | A | 80 | Claude Code |
| orbit-notionnexu-io/open-design | Skill | 106k | yesterday | Apache-2.0 | A | 117 | Claude Code, Codex |
| agentmailpaperclipai/paperclip | Skill | 111k | yesterday | MIT | A | 30 | Claude Code, Codex |
| Cortexdanielmiessler/LifeOS | Skill | 20k | yesterday | MIT | A | 184 | Claude Code |
| rowboat-spacesrowboatlabs/rowboat | Skill | 18k | yesterday | Apache-2.0 | A | 60 | Claude Code, Codex |
| feishuHmbown/CodeWhale | Skill | 41k | 5d ago | MIT | A | 33 | Claude Code, Codex |
| pinchtab-mcppinchtab/pinchtab | Skill | 11k | 10d ago | MIT | A | 52 | Claude Code, Codex |