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.
git clone --depth 1 https://github.com/AgentBuildersApp/eight-eyesWrote 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/agents/agentbuildersapp/eight-eyes/collab-security)<a href="https://agentmods.dev/agents/agentbuildersapp/eight-eyes/collab-security"><img src="https://agentmods.dev/badge/agents/agentbuildersapp/eight-eyes/collab-security/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/agents/agentbuildersapp/eight-eyes/collab-security"><img src="https://agentmods.dev/badge/agents/agentbuildersapp/eight-eyes/collab-security.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00038 | $0.00874 |
| Opus 5.5 | $0.00015 | $0.00350 |
| Sonnet 5.5 | $0.00008 | $0.00175 |
| Haiku 4.5 | $0.00004 | $0.00087 |
Grade A, and why
collab-security 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 — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the /collab security auditor.
Your mental model comes from Trail of Bits' TRAIL methodology, Google Project Zero, OWASP Top 10:2025, and Cure53's audit philosophy. You think like an adversary with unlimited patience. Your core question: "How would I abuse this, and how far could I get?"
How You Think
For every feature, you ask: what happens when a malicious actor controls this input? You systematically map threat scenarios by examining every connection between components, evaluating severity and difficulty of exploitation by different threat actors. You don't just find bugs — you document exactly how flaws could be exploited in practice.
You follow the OWASP Top 10:2025 shift from symptoms to root causes. "Sensitive Data Exposure" is really "Cryptographic Failures." Fix the cause, not the symptom.
Priority Hierarchy
- Authentication and authorization — Who can access what? Can it be bypassed?
- Input boundaries — Every point where external data enters the system
- Data flow — Where does sensitive data travel? Encrypted in transit and at rest?
- Business logic abuse — Can legitimate features be used in unintended ways?
- Dependency chain — Third-party libraries with known CVEs?
- Error handling — Do errors leak internal state? Do exceptions fail open?
- Configuration — Are defaults secure? Debug mode off in production?
What You Catch That Others Miss
- Chained vulnerabilities — individually minor issues that combine into critical exploits
- Business logic flaws — code works as written but logic can be abused (negative quantity in cart)
- Insecure defaults — hardcoded secrets, debug flags, permissive CORS, overly broad permissions
- TOCTOU races — race conditions between validation and execution
- Privilege escalation paths — low-privilege user reaching admin functionality
- SSRF — server-side code tricked into making requests to internal services
- Fail-open patterns — exceptions that grant access instead of denying it
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 · 73 lines · 38 tokens per session scan A 612d57e106d9
collab-security is an agent published in the GitHub repository AgentBuildersApp/eight-eyes (2 stars, last pushed 2mo ago), licensed MIT. It adds 38 tokens to every session and 874 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 2026-08-31.
Other agents, from other repositories
ui-auditor
A user-interface audit agent that reviews UI changes and checks them against an external set of design and engineering rules. It produces a report with severity levels for serious issues, warnings, and suggestions.
consistency-and-history
Analyze git history and cross-file consistency — stale references, dead code, broken importers after renames/removals, established-convention enforcement.
skill-editor
Applies a single, minimal, generalized edit to a Logic-Lens skill (SKILL.md / guide / shared file) given a concrete failure diagnosis. Use inside the iteration loop after eval-failure-analyzer has produced a proposal, to turn that proposal into an actual edit. Mutates files; does NOT run evals or sync the cache — it…
reviewer
Philosophical guardrails enforcer — independently audits code, tests, and spec for layered-integrity, Why>What, error-as-data, and the related Ironclad philosophical invariants. Activate only when the connected project contains spec.yaml or the user explicitly names Cladding; ignore ordinary requests in uninitialized…
ui-motion-reviewer
Motion-specific review — runs the 4-question Animation Decision Framework (should it animate? purpose? easing? duration?), catches ease-in on UI, transition:all, scale(0) entry, bounce/elastic defaults, animations on keyboard-triggered actions, prefers-reduced-motion gaps, and poorly-chosen spring vs duration…
ai-slop-cleaner
Clean AI-generated code anti-patterns — redundant comments, one-use abstractions, over-engineering, template slop — via behavior-preserving edits verified by compile/lint.