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 commands/fprochazka/claude-code-plugins/wrap-upgit clone --depth 1 https://github.com/fprochazka/claude-code-pluginsWhat 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 | $0.00024 | $0.01532 |
| Opus 5 | $0.00012 | $0.00766 |
| Sonnet 5 | $0.00005 | $0.00306 |
| Haiku 4.5 | $0.00002 | $0.00153 |
Grade A, and why
wrap-up 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 — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Wrap Up
Close out finished work: post one dense comment to the ticket with everything this session learned that is worth keeping, then mark the ticket completed in the issue tracker.
Scope
$ARGUMENTS
Sources
You (the orchestrator) write the comment yourself — never a subagent. The primary source is the conversation you are holding, and no subagent can see it. Priority order: (1) the session itself, including every production check, query result, and dashboard reading done in it, (2) ./.claude/plans/*.md for what the work set out to do, (3) the diff and the MR discussion, as far as the session already read them. Do not go exploring for new material and do not re-run the analysis — a fact the session never established stays out of the comment.
Numbers must be the real ones you saw. Never reconstruct a figure from memory of a shape ("about 5%"). If you cannot state a number exactly as it was measured, drop it or state the reading you actually have.
Process
- Take the scope from the session. This command normally runs at the end of the work it wraps up, so the ticket, the MR, and the checks are already in the conversation. Use them. The scope above only narrows or overrides that, and is empty most of the time. Fall back to the branch name or the MR title when the session is genuinely fresh, and ask the user only when the session covered several tickets and you cannot tell which one is finished. Load the skill for the issue tracker in play — you need it to post and to close.
- Judge from the session whether the work is finished — MR merged, CI green on the target branch, every step of the plan done. Do not go and look any of it up. You either watched it happen or you did not. If the session left something open, say so in your reply to the user and ask before you mark the ticket completed. If the session never established a state at all, treat it as unknown, keep it out of the comment, and mention it to the user. Post the comment either way.
- Draft the comment to
<scratchpad>/<ticket-id>-wrap-up.md. Follow the structure and the density rules below. When the material calls for more than one comment, draft each into its own numbered file —<ticket-id>-wrap-up-1-outcome.md,-2-verification.md, and so on. - Show the user a high-level summary and the file paths. They can read the drafts from the files. Iterate if they want changes.
- Post the comments by reading each file into the argument (
"$(cat <path>)"), not by inlining the text. Post them in file order, so the ticket reads top to bottom. - Mark the ticket completed — read the tracker's real workflow states and pick its terminal one instead of guessing a name. Do not change assignee, labels, project, or priority.
- Report — ticket URL, the state you set, and the follow-ups worth filing. Point at
/sdlc:ticket-newfor those. Do not file them yourself.
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 · 65 lines · 24 tokens per session scan A 595a4fc2e565
wrap-up is a command published in the GitHub repository fprochazka/claude-code-plugins (11 stars, last pushed 4d ago), licensed MIT. It adds 24 tokens to every session and 1,532 once invoked, about $0.0001 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 commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.