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 instructions/rjmurillo/ai-agents/securitygit clone --depth 1 https://github.com/rjmurillo/ai-agentsWrote 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/instructions/rjmurillo/ai-agents/security)<a href="https://agentmods.dev/instructions/rjmurillo/ai-agents/security"><img src="https://agentmods.dev/badge/instructions/rjmurillo/ai-agents/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 | $0.00832 | $0.00832 |
| Opus 5 | $0.00416 | $0.00416 |
| Sonnet 5 | $0.00166 | $0.00166 |
| Haiku 4.5 | $0.00083 | $0.00083 |
Grade A, and why
ai-agents security.instructions.md 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 today.
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 — 42 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Security File Rules
These paths hold threat models, benchmarks, workflows, and hooks that protect the supply chain. Changes need evidence, not opinion.
MUST
- Security agent review. Changes MUST be reviewed by the
securityagent (or a human with equivalent authority) before merge. Security review is always-on and cannot be skipped. - Evidence-based. Every change MUST cite a CVE, CWE, OWASP reference, ADR, or documented threat. No speculative hardening.
- No secrets. MUST NOT commit credentials, tokens, signing keys, or API keys. Use the configured secrets manager.
- Pin Actions to SHA. Workflow changes MUST pin every third-party Action to a commit SHA. Floating tags (
@v4,@main) are prohibited. - Threat model updates. Changes that introduce new attack surface MUST update the relevant benchmark under
.agents/security/benchmarks/or cite why no update is required. - Test coverage. Security-critical code MUST have 100% coverage per
AGENTS.mdstandards. - Eureka: gate availability is gate correctness. A required security review that does not run MUST produce a blocking verdict. Infrastructure failure is not a security pass. Issue #4777 records the failure where ten missing reviews produced a green required check.
SHOULD
- Run security scan locally. SHOULD run the
security-scanskill (.claude/skills/security-scan/scripts/scan_vulnerabilities.py) before pushing. The internal scanner detects CWE-78 (command injection) only; CWE-22 (path traversal) detection is delegated to CodeQL in CI per ADR-054 amendment 2026-05-02. - Use the
security-detectionskill. SHOULD detect security-relevant file changes via the skill and route to the security agent. - Threat modeling. SHOULD use the
threat-modelingskill (OWASP STRIDE) for non-trivial changes.
MUST NOT
- MUST NOT lower severity thresholds in
SECURITY-SEVERITY-CRITERIA.mdwithout governance ADR. - MUST NOT skip security checks in CI.
- MUST NOT merge security-sensitive changes without explicit approval, even when auto-merge labels are applied.
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.
- today First seen · 42 lines · 832 tokens per session scan A 00d4bcfbb533
ai-agents security.instructions.md is an instructions file published in the GitHub repository rjmurillo/ai-agents (45 stars, last pushed today), licensed MIT. It adds 832 tokens to every session, about $0.0042 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 2026-09-03.
Other instructions, from other repositories
claude-capsule-kit CLAUDE.md
Instructions for arpitnath/claude-capsule-kit, covering capsule kit v4.0, rules, context system (capsule), wikilink layer and docs.
carmenta AGENTS.md
AGENTS.md instructions for carmentacollective/carmenta, covering project context for ai assistants, anti-patterns to avoid, always apply rules, project overview and directory documentation.
carmenta CLAUDE.md
Claude Code instructions for carmentacollective/carmenta, a project described as: A heart-centered AI for creating at the speed of thought.
Alex_Plug_In automated-quality-gates.instructions.md
Automated pre-publish quality gates — preventing regressions through build pipeline enforcement.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.