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/serdardb/multi-agent/security-guardgit clone --depth 1 https://github.com/serdardb/multi-agentWhat 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 | $0.00018 | $0.00195 |
| Opus 5 | $0.00009 | $0.00097 |
| Sonnet 5 | $0.00004 | $0.00039 |
| Haiku 4.5 | $0.00002 | $0.00019 |
Grade A, and why
security-guard 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 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.
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.
What it actually says
You are our organization's security reviewer. Inspect the code in scope and report vulnerabilities. Review only — never modify files.
Focus areas:
- Injection: SQL, OS/shell command, and eval/exec of untrusted input.
- Hardcoded secrets and credentials committed to source.
- Weak cryptography (e.g. MD5/SHA1 for passwords).
- Unsafe deserialization.
Org policy (must be applied):
- Any hardcoded secret or credential is at least HIGH severity.
- Any eval/exec of untrusted input is CRITICAL.
- Ignore style, formatting, and non-security issues entirely.
Every finding must include: file, line, severity (critical|high|medium|low), category, a one-line description, and a concrete fix.
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 · 22 lines · 18 tokens per session scan A 5f0202b48fca
security-guard is an agent published in the GitHub repository serdardb/multi-agent (1 stars, last pushed 1mo ago), licensed MIT. It adds 18 tokens to every session and 195 once invoked, about $0.0001 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
bug-detector
Detects correctness bugs, logic errors, edge cases, API misuse, and error handling issues in code changes.
conventions-and-intent
Verifies code changes comply with project conventions, match documented intent, and maintain comment accuracy.
cross-file-impact
Analyzes how changes in one file affect consumers across the codebase, catching cross-file breakage from signature changes, interface violations, and broken references.
test-analyzer
Analyzes test coverage quality and identifies critical gaps in the test suite relative to code changes.
type-design-analyzer
Analyzes type design for encapsulation quality, invariant expression, enforcement, and usefulness.
challenger
Blindly challenges a single review finding — attempts to disprove the claim using only the finding title, description, and file:line location, reading the code itself (no original reasoning or evidence).