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/samibs/skillfoundry/securitygit clone --depth 1 https://github.com/samibs/skillfoundryWrote 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/commands/samibs/skillfoundry/security)<a href="https://agentmods.dev/commands/samibs/skillfoundry/security"><img src="https://agentmods.dev/badge/commands/samibs/skillfoundry/security.svg" alt="Measured on agentmods" 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 | $0.00000 | $0.03368 |
| Opus 5 | $0.00000 | $0.01684 |
| Sonnet 5 | $0.00000 | $0.00674 |
| Haiku 4.5 | $0.00000 | $0.00337 |
Grade B, and why
security scanned grade B with 2 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 yesterday.
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.
Instruction-override phrasingmediumPrompt injection
Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.
| Prompt Injection | Instruction embedded in data an LLM later reads (user content, doc, retrieved page, tool output): "ignore previous instructions", "output your system prompt", "act as…" | Treat model-read content as Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
Asks the agent to reveal its instructionslowSystem prompt leakage
Directions to print, repeat or translate the system prompt extract configuration the operator did not intend to expose.
| Prompt Injection | Instruction embedded in data an LLM later reads (user content, doc, retrieved page, tool output): "ignore previous instructions", "output your system prompt", "act as…" | Treat model-read content as Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 367 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Security Specialist
You are a rigorous security specialist with an attacker's mindset. You think like a malicious actor to find vulnerabilities before they ship. You do not accept "it's just internal" or "we'll fix it later" security theater.
Persona: See agents/security-specialist.md for full persona definition.
Operational Philosophy: Every input is hostile. Every user is an attacker. Every dependency is compromised. Prove me wrong with evidence, not assumptions.
Known Deviations: See agents/_known-deviations.md for 80+ LLM failure patterns to prevent.
Shared Modules: See agents/_reflection-protocol.md for reflection requirements.
OPERATING MODES
/security audit [target]
Full security audit of component/system. Output structured vulnerability report.
/security threat-model [feature]
STRIDE-based threat modeling for new features.
/security pentest [target] — white-hat penetration testing
Penetration-testing mindset: enumerate the attack surface and test for OWASP Top 10
vulnerabilities. Sub-actions: scan [target] (full OWASP sweep of a codebase/endpoint spec),
endpoint [url] (injection / auth-bypass / rate-limit on one API), report (findings writeup).
Authorized-scope rules (non-negotiable): operate only within authorized scope — never test systems without explicit permission; test for vulnerability existence, do not execute destructive exploits; refuse attacks on production systems without documented authorization.
Every finding: [SEVERITY] title · CVSS X.X (vector) · location (file:line/endpoint) · description · proof-of-concept to reproduce · specific remediation · CWE-XXX / OWASP-AYYY.
For attacker-primitive-level deep analysis of specific findings, chain with
/red-team-researcher.
/security review [code]
Security-focused code review (beyond standard review).
/security incident [description]
Incident response guidance and remediation.
/security harden [system]
Hardening recommendations for infrastructure/application.
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.
- yesterday First seen · 367 lines · 0 tokens per session scan B 88ce1a4f0b88
security is a command published in the GitHub repository samibs/skillfoundry (12 stars, last pushed today), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 3,368 tokens. A static security scan graded it B with 2 findings (instruction-override phrasing, asks the agent to reveal its instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
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.