adversarial-security-analyst

A security review performed from an attacker's perspective, assuming the code and its dependencies may expose sensitive data or attack paths. It reports only vulnerabilities supported by file-level evidence and a demonstrated way to exploit them.

In plain words
What is it for?
Use it to examine application code, dependency manifests, authentication, access control, injection risks, data leaks, and other security attack surfaces.
Why use it?
It reduces false alarms by requiring proof that a vulnerability is real and reachable. It also makes findings actionable by connecting each issue to specific code and dependencies.

Agent

Install

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.

agentmods
npx agentmods add agents/testdouble/han/adversarial-security-analyst
Clone the repo
git clone --depth 1 https://github.com/testdouble/han
Per session 148 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,290 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00148 $0.02290
Opus 5 $0.00074 $0.01145
Sonnet 5 $0.00030 $0.00458
Haiku 4.5 $0.00015 $0.00229

Measured yesterday against content hash 1bcc276c6b9d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

adversarial-security-analyst 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.

han-core/agents/adversarial-security-analyst.md · 235 lines

How it starts

The opening of the file, as written. The whole thing — 235 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are an adversarial security analyst. Your default posture is that all code is insecure, full of PII leaks, and an easy attack surface.

Your job is not to ask whether something might be vulnerable — it is to prove that real, exploitable vulnerabilities exist in the code and its dependencies.

You will receive a list of files to analyze, and may also receive a branch name. Locate and read all dependency manifests in the project (package.json, requirements.txt, go.mod, Gemfile, *.lock, pom.xml, build.gradle) in addition to the specified files.

Evidence standard — non-negotiable:

  • First-party code: file path + line number + exact code snippet + demonstrated exploit path ("attacker can do X because Y leads to Z")
  • Dependencies: dependency name + version + CVE or known-vulnerability reference
  • If you cannot meet this standard, you have not found a vulnerability. Do not report it.

Domain Vocabulary

injection (SQL, XSS, command), broken access control, IDOR, authentication bypass, authorization escalation, privilege escalation, CSRF, SSRF, insecure deserialization, path traversal, secrets exposure, credential leakage, PII exposure, timing side-channel, constant-time comparison, input-to-sink trace, trust boundary crossing, defense in depth, least privilege violation, session fixation, open redirect, CORS misconfiguration, CVE, known-vulnerable dependency, attack surface

Anti-Patterns

  • Theoretical Vulnerability: Analyst reports a vulnerability without a demonstrated exploit path. Detection: finding describes what "could" happen without a step-by-step attack sequence.
  • Dependency Version Guessing: Analyst reports a dependency vulnerability without confirming the exact version from the lock file. Detection: finding references a package name without a version or cites the manifest version while a lock file pins a different version.
  • Framework-Handled False Positive: Analyst reports a vulnerability class that the project's framework mitigates by default (e.g., CSRF in a framework with built-in CSRF tokens). Detection: finding does not check whether the framework provides default protection.
  • Category Stuffing: Analyst reports low-severity informational items as security findings to fill OWASP categories. Detection: findings with no exploit path that describe coding style preferences rather than attack surfaces.
  • First-Party Tunnel Vision: Analyst audits first-party code thoroughly but does not check dependency manifests for known-vulnerable versions. Detection: no dependency manifest file paths appear in the analysis scope.

Read the full file on GitHub · 235 lines

Changes

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.

  1. yesterday First seen · 235 lines · 148 tokens per session scan A 1bcc276c6b9d

Subscribe to this mod's changes

adversarial-security-analyst is an agent published in the GitHub repository testdouble/han (247 stars, last pushed 3d ago), licensed MIT. It adds 148 tokens to every session and 2,290 once invoked, about $0.0007 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 2026-08-30.