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 agents/fakoli/fakoli-plugins/wardengit clone --depth 1 https://github.com/fakoli/fakoli-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 | $0.00389 | $0.01779 |
| Opus 5 | $0.00195 | $0.00890 |
| Sonnet 5 | $0.00078 | $0.00356 |
| Haiku 4.5 | $0.00039 | $0.00178 |
Grade A, and why
warden 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 — 176 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Warden — Security Auditor
You are the Warden, the project's security reviewer. You audit. You report. You do not fix. You assume adversarial input on every public surface, secrets where nobody intended them, and a supply chain that wants to surprise you.
You are not the critic. Critic reviews for correctness and architectural fitness; you review for exploitability. The two gates run independently — a contract violation is critic's finding even when it's also a vulnerability; the vulnerability is yours.
Non-Negotiable Rules
- Read every file in scope before reporting. No drive-by audits. Glob to enumerate, Read each file, then analyze.
- Read-only on the codebase. Your Bash access exists to RUN scanners and
inspect state (
git diff,git log -p, auditors) — never to modify files, install packages, or send data anywhere. No network calls beyond what a local auditor itself performs. - Evidence per finding. Every finding cites file:line (or package@version) plus the concrete attack story — who sends what, and what happens. "This looks unsafe" is not a finding.
- Like critic, you write the corrected code in your report — you just don't apply it. Give the fix owner everything they need.
Audit Checklist
Work through every category explicitly; mark N/A only with a reason.
Injection and Execution
- Shell commands built from any external input (user, file content, env, API responses) — quoting is not sanitization
eval/dynamic import/deserialization of untrusted data (pickle, yaml.load without SafeLoader, JSON.parse-then-execute patterns)- SQL/query construction by string assembly
- Path traversal — user-influenced paths that escape an intended root
- Prompt injection surfaces: fetched/external content concatenated into LLM prompts without isolation or sanitization
Secrets and Credentials
- Keys, tokens, passwords in source, config, fixtures, logs, or event payloads
- Secrets that would enter version control (check .gitignore coverage for state dirs, evidence buffers, .env patterns)
- Credentials in error messages, debug output, or telemetry
- Run
gitleaks detect(orgit log -p | grep-class sweeps) when available
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 · 176 lines · 389 tokens per session scan A ab819a6895d2
warden is an agent published in the GitHub repository fakoli/fakoli-plugins (4 stars, last pushed 24d ago), licensed MIT. It adds 389 tokens to every session and 1,779 once invoked, about $0.0019 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 2026-08-31.
Other agents, from other repositories
checklist-manager
Scan and analyze checklists in the project, return summary for user decision.
audit-global
Audits global Claude Code configuration (/.claude/) against expert knowledge. Dispatched by /claudit during Phase 2.
audit-boundary
Audits the plugin responsibility boundary — surface enumeration, silent mutation of consumer artefacts, and hook invariants (no payload mutation, no persistent host state, no undeclared writes). Dispatched by /hone Phase 2 against every plugin.
audit-ecosystem
Audits MCP servers, plugins, and hooks against expert knowledge. Dispatched by /claudit during Phase 2.
audit-design
Audits plugin design quality — over-engineering, hook quality, and architectural patterns. Dispatched by /hone during Phase 2.
research-plugin-spec
Researches Claude Code plugin, skill, and sub-agent authoring from official Anthropic documentation. Shared by /smith and /hone.