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/ByamB4/find-cve-agentWrote 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/byamb4/find-cve-agent/recon)<a href="https://agentmods.dev/commands/byamb4/find-cve-agent/recon"><img src="https://agentmods.dev/badge/commands/byamb4/find-cve-agent/recon.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 5.1 | $0.00040 | $0.00793 |
| Opus 5 | $0.00020 | $0.00396 |
| Sonnet 5 | $0.00008 | $0.00159 |
| Haiku 4.5 | $0.00004 | $0.00079 |
Grade A, and why
recon scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -s "https://api.npmjs.org/downloads/point/last-week/<package>" How it starts
The opening of the file, as written. The whole thing — 99 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/recon
Find promising targets in a specific package category for CVE hunting.
Process
Step 1: Search Package Registries
Based on the category argument, search for relevant packages:
# npm search
npm search "$ARGUMENTS" --long 2>/dev/null | head -20
# GitHub search (adjust language and keywords per category)
gh search repos "$ARGUMENTS" --language javascript --stars 500..15000 --sort stars --limit 20
gh search repos "$ARGUMENTS" --language python --stars 500..15000 --sort stars --limit 20
gh search repos "$ARGUMENTS" --language go --stars 500..15000 --sort stars --limit 20
Step 2: Filter Candidates
For each result, quickly evaluate:
- Stars: 500-15K? (skip if outside range)
- Last commit: within 6 months? (skip if abandoned)
- Already in REGISTRY.md? (skip if yes)
- More than 10 existing CVEs? (skip if over-audited)
Step 3: Deep Evaluation (Top 5)
For the top 5 candidates, gather detailed info:
# Download counts
curl -s "https://api.npmjs.org/downloads/point/last-week/<package>"
# Repo details
gh repo view <owner>/<repo> --json stargazerCount,pushedAt,description,licenseInfo
# Security advisories count
gh api "repos/<owner>/<repo>/security-advisories" 2>/dev/null | python3 -c "import sys,json; data=json.load(sys.stdin); print(len(data))" 2>/dev/null || echo "0"
Step 4: Rank and Present
Present candidates to the Director ranked by promise:
RECON RESULTS: <category>
#1. <package-name>
Stars: <count> | Downloads: <count>/week | Language: <lang>
Attack surface: <brief>
Existing CVEs: <count>
Top vector: <most promising>
Rating: HIGH / MEDIUM / LOW
#2. <package-name>
...
#3. <package-name>
...
Step 5: Propose Top Pick
Recommend the best candidate to the Director:
Proposed target: <package-name>
Reason: <why this is the best candidate>
Approve? (I'll create the full brief on approval)
Category Keywords
| Category Argument | Search Terms |
|---|---|
| csv-parsers | csv parse, csv parser, csv reader |
| xml-parsers | xml parse, sax parser, xml reader |
| template-engines | template engine, mustache, handlebars, nunjucks |
| archive-libs | zip extract, tar extract, archive, decompress |
| validators | validate, schema validator, json schema |
| file-handlers | file upload, multipart, file process |
| serializers | serialize, deep clone, deep merge, marshal |
| url-parsers | url parse, uri, query string |
| image-libs | image process, sharp, jimp, image resize |
| pdf-libs | pdf parse, pdf generate, pdf reader |
| markdown | markdown parse, markdown render, markdown-it |
| yaml | yaml parse, yaml load |
| auth-libs | jwt, oauth, session, authentication |
| crypto | encrypt, decrypt, hash, cipher |
| cli-tools | cli, command line, terminal, exec |
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.
- 8d ago First seen · 99 lines · 40 tokens per session scan A a3ccc7ab4ba4
recon is a command published in the GitHub repository ByamB4/find-cve-agent (48 stars, last pushed 5mo ago), licensed Apache-2.0. It adds 40 tokens to every session and 793 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other commands, from other repositories
learn
Force claude-smart to extract learnings from this session now.
init
Install the formatters this repository needs, with every command visible before it runs.
brand-generate
Generate an on-brand document from a saved Brand Profile.
genshijin-compress
A command for safely shortening Markdown or text files into the genshijin style.
check-dev
Type-check a Z specification with fuzz.
adr-supersede
Create a new ADR that supersedes an existing one.