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 SethGammon/Citadel --skill unharnessgit clone --depth 1 https://github.com/SethGammon/CitadelWrote 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/sethgammon/citadel/unharness)<a href="https://agentmods.dev/skills/sethgammon/citadel/unharness"><img src="https://agentmods.dev/badge/skills/sethgammon/citadel/unharness/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/sethgammon/citadel/unharness"><img src="https://agentmods.dev/badge/skills/sethgammon/citadel/unharness.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00059 | $0.01008 |
| Opus 5.5 | $0.00024 | $0.00403 |
| Sonnet 5.5 | $0.00012 | $0.00202 |
| Haiku 4.5 | $0.00006 | $0.00101 |
Grade A, and why
unharness 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 — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/unharness - Receipt-owned exit from Citadel
Orientation
Use this skill when the user wants Citadel removed from a project. Exit is a two-step, receipt-owned operation. A plan never mutates the project. Apply requires the exact saved plan and its confirmation token.
Do not infer ownership from familiar paths. Do not delete .planning/,
.citadel/, runtime settings, hooks, skills, or registrations merely because
they look Citadel-related.
Protocol
1. Locate Citadel and inspect adoption authority
Read .citadel/plugin-root.txt when present. Otherwise use the directory
containing this skill. Check for .citadel/adoption/active.json; the adoption
planner also checks the private recovery ledger.
2. Create a no-write leave plan
Choose a plan path inside the project, such as
.planning/adoption/leave.plan.json, and run:
node {citadelRoot}/scripts/adopt.js leave plan \
--target {projectRoot} \
--out {planPath} \
--json
Show the user:
- exact files that will be removed or restored;
- modified or ambiguous footprint entries that will be retained;
- portable archive path;
- runtime registrations whose removal evidence is
unknown; - the plan digest and confirmation token.
If the result is NOT_ADOPTED, stop. Create a conservative legacy inventory:
node {citadelRoot}/scripts/adopt.js import plan {citadelRoot} \
--target {projectRoot} \
--out {importPlanPath} \
--json
Apply that import only after its own explicit approval. Then create a new leave plan from the resulting receipt. Never describe legacy cleanup as exact removal.
3. Obtain explicit approval
Leaving is destructive. Ask the user to approve the displayed plan and exact confirmation token. Do not apply from an unsaved or regenerated plan.
4. Apply and verify
After approval:
node {citadelRoot}/scripts/adopt.js leave apply {planPath} \
--confirm {confirmationToken} \
--json
node {citadelRoot}/scripts/adopt.js doctor \
--target {projectRoot} \
--json
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 · 138 lines · 59 tokens per session scan A 974d435ec086
unharness is a skill published in the GitHub repository SethGammon/Citadel (927 stars, last pushed yesterday), licensed MIT. It adds 59 tokens to every session and 1,008 once invoked, about $0.0002 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 2026-08-30.
Other skills, from other repositories
review
Get a person's review of an artifact you produced or were handed: a plan, spec, report or other Markdown file. Use when you have a file that someone should read and react to before work continues, or when the user asks to send something for review. Picks the Terminal (TUI) or the Web from this project's recorded…
fo-status-viewer
First-officer status query/mutate/display surface — the status command flag docs, --set field docs, canonical captain-facing invocations, the Captain-Facing State Display rendering, and the GitHub-issue-filing approval gate. Invoke at the first ad-hoc status question, --set mutation, --next-id/--resolve lookup, or…
drift-detect
Use when the user says "plan drift", asks whether the roadmap, plans, or docs still match the code, or is deciding what to rebuild when restarting a stalled project. For doc-vs-code drift inside a specific diff, use sync-docs.
github-backlog-triage
Use when the user invokes backlog triage for a GitHub repo's open issues and PRs. Not for proactive triage, non-GitHub trackers, or single bug issues: use github-bug-report-triage.
gitlab-ops
Use this skill when performing VCS operations on GitLab or GitHub repositories — creating, updating, or closing issues and MRs, applying label taxonomy, running glab/gh CLI commands, or resolving project paths dynamically. Acts as the single source of truth for CLI command syntax and label conventions; consuming…
plan
Use this skill when performing structured project planning and PRD generation with three modes: new (project kickoff with repo scaffolding), feature (compact feature PRD), retro (data-driven retrospective). All modes share a researched Q&A engine that dispatches parallel Explore agents before each question wave…