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/ashkaan/contextium/closenpx skills add Ashkaan/contextium --skill closegit clone --depth 1 https://github.com/Ashkaan/contextiumWhat 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 | $0.00121 | $0.03362 |
| Opus 5 | $0.00060 | $0.01681 |
| Sonnet 5 | $0.00024 | $0.00672 |
| Haiku 4.5 | $0.00012 | $0.00336 |
Grade A, and why
close 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 — 242 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/close — Session Wrap
The "wrap" verb of the loop. A session isn't closed by git commit alone — the journal entry is
load-bearing. /close writes the journal, updates any touched project README, commits, and pushes, in
one pass. For substantial changes that weren't already reviewed by /implement, it runs
/implement-audit first so blind spots get caught before the work is sealed.
/close is also the auto-tail of the producer verbs. /spec and /implement auto-invoke /close
on clean completion per references/auto-close-gate.md — so in the
normal loop nobody types /close; the producing verb fires it. When auto-invoked, the close-fired
session marker (scripts/close-fired.sh) guards against a double-fire if a held question resumes the
verb after /close had already started. Typing /close yourself still works for ad-hoc sessions that
didn't run a producer verb.
Critical
- The journal is required. Do not stop after
git commit. The journal entry (§2) is what makes a past decision reconstructable later per@rule:journal-format; the commit alone is not "closed." - Audit substantial ad-hoc changes — and only those. §1 asks a marker file whether this session
was already audited. If it was, the review does not run again and its trailer rides onto the commit
instead. If it wasn't, and the session changed real code,
/implement-auditruns here and every ready finding is fixed before the commit. Do not skip it to close faster. - Land the full scope. Fix the audit findings this session per
@rule:no-deferral; don't stage them in a README as "follow-ups."
code-review — pre-close audit (substantial ad-hoc changes only)
/implement-audit is the loop's single code reviewer, normally fired by /implement. /close is the
backstop for substantial work done OUTSIDE /implement — a quick fix that grew, a refactor you did by
hand.
Ask the marker, don't judge it:
bash .claude/skills/implement-audit/scripts/audit-dedupe.sh status
What ships with it
13 files 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.
- references/auto-close-gate.md 4.4 KB
- scripts/close-fired.sh 4.8 KB runs code
- scripts/close-fired.test.sh 6.4 KB runs code
- scripts/detect-mode.sh 1.6 KB runs code
- scripts/detect-mode.test.sh 3.3 KB runs code
- scripts/next-implement-command.sh 5.9 KB runs code
- scripts/next-implement-command.test.sh 3.8 KB runs code
- scripts/project-remaining-work.sh 9.3 KB runs code
- scripts/project-remaining-work.test.sh 9.6 KB runs code
- scripts/push-with-retry.sh 2.6 KB runs code
- scripts/push-with-retry.test.sh 3.3 KB runs code
- scripts/safe-commit.sh 4.0 KB runs code
- scripts/safe-commit.test.sh 7.0 KB runs code
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 · 242 lines · 121 tokens per session scan A c2b9693e8bc0
close is a skill published in the GitHub repository Ashkaan/contextium (32 stars, last pushed 5d ago), licensed MIT. It adds 121 tokens to every session and 3,362 once invoked, about $0.0006 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
micro-commit
Enforces strict, atomic commit hygiene. Splits large diffs into logical, revertible units.
git-doc-updater
Git 提交前文檔同步。觸發：docs、文檔、sync docs、發布。.
git-precommit
提交前編排器 + Paper-Aware Pre-Commit Hooks。 LOAD THIS SKILL WHEN: commit、提交、推送、做完了、收工 CAPABILITIES: 記憶同步、文檔更新、Paper 品質把關、Git 操作.
pr-draft-summary
Create the required PR-ready summary block, branch suggestion, title, and draft description for openai-agents-python. Use before the final response whenever the current task changed runtime code, tests, examples, build/test configuration, or docs with behavior impact, regardless of perceived change size and including…
git-commit
Generate well-formatted git commit messages following conventional commit standards.
agentic-code-orchestrator
Unified codebase manipulation, AI deployment, data analysis, and academic delivery engine. Absorbs 6 coding protocols + data-analysis + academic-delivery + spec-driven-dev.