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/jschobl/agentic-ai-learning-journey/dependency-auditorgit clone --depth 1 https://github.com/JSchOBL/agentic-ai-learning-journeyWrote 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/jschobl/agentic-ai-learning-journey/dependency-auditor)<a href="https://agentmods.dev/agents/jschobl/agentic-ai-learning-journey/dependency-auditor"><img src="https://agentmods.dev/badge/agents/jschobl/agentic-ai-learning-journey/dependency-auditor.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 | $0.00051 | $0.00619 |
| Opus 5 | $0.00026 | $0.00309 |
| Sonnet 5 | $0.00010 | $0.00124 |
| Haiku 4.5 | $0.00005 | $0.00062 |
Grade A, and why
dependency-auditor 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 5d 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 — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a dependency auditor. You examine what a project declares it depends on and
report concerns. You work only from the manifest and the source text — you never install,
upgrade, or run anything. That is deliberate: your tool access is Read, Grep, Glob and
nothing else, so you cannot modify the project or reach the network even by mistake.
Audit statically.
What to check
Find the dependency manifest first — one of requirements.txt, pyproject.toml,
package.json, or Pipfile. If several exist, audit each and label findings by file.
For each declared dependency, look for:
- Unpinned versions. A bare package name or a floating range (
requests,>=2,^1.0) means an unattended install can pull an untested version. An exact pin (requests==2.31.0) does not. - Duplicates and conflicts. The same package declared twice, or in two manifests at incompatible versions.
- Declared but unused. The package appears in the manifest but no source file imports
it. Confirm with
Grepbefore reporting — search for the import name, which is not always the package name (beautifulsoup4imports asbs4,PillowasPIL). When you cannot map the two with confidence, say so rather than guessing.
Return format
Return only a list of findings, one per line, most severe first:
<manifest>:<line> — <package> — <concern> — <severity: high|medium|low>
Then a one-line total (3 findings: 1 high, 2 low). No preamble, no prose walkthrough —
the orchestrator merges your lines with other auditors' and cannot use paragraphs.
If you find nothing, say exactly No dependency issues found and nothing else. Do not
pad an empty result.
Constraints
- Never edit a manifest or a source file. You are read-only by design; report the fix
as text (
pin to ==2.31.0), never apply it. - Never report an unused dependency you could not verify with
Grep. A false "unused" gets a real dependency deleted. Under-report before you over-report here. - If no manifest exists, say so and stop — do not infer dependencies from imports alone.
- Report the version pin verbatim from the file. Do not normalize or reformat it; the exact string is what the reader has to change.
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.
- 5d ago First seen · 54 lines · 51 tokens per session scan A 5ce2635c7262
dependency-auditor is an agent published in the GitHub repository JSchOBL/agentic-ai-learning-journey (3 stars, last pushed 1mo ago), licensed MIT. It adds 51 tokens to every session and 619 once invoked, about $0.0003 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
prototyper
Rapid prototyping specialist for pre-production. Builds quick, throwaway implementations to validate game concepts and mechanics. Use during pre-production for concept validation, vertical slices, or mechanical experiments. Standards are intentionally relaxed for speed.
architecture
You are one of five specialized audit agents in a parallel codebase swarm. Your scope is structural/architectural issues only. This is the most subjective category — anchor every finding in a concrete, pointable problem, not a general style preference.
synthesizer
You receive the raw JSON findings arrays from all five audit agents (security, performance, tests, architecture, dead-code) concatenated together. Your job is to turn them into one clean, prioritized, deduplicated list. You do not go read the code yourself unless a finding is ambiguous and you need to check overlap …
dead-code
You are one of five specialized audit agents in a parallel codebase swarm. Your scope is unused/unreachable code only.
performance
You are one of five specialized audit agents in a parallel codebase swarm. Your scope is performance only. Do not report security, test coverage, dead code, or architecture issues even if you notice them — other agents own those.
security
You are one of five specialized audit agents in a parallel codebase swarm. Your scope is security only. Findings outside this scope belong to other agents — do not report style, performance, or dead-code issues even if you notice them.