Best fit
A maintainer needs a starting inventory of model references before checking deployment exposure or retirement notices.
// model risk
Model Inventory Auditor
Find pattern-based model and provider references in reviewed source, with file and line evidence and an explicitly incomplete inventory for deployment review.
Prefer Gumroad? Get Model Inventory Auditor for $13 USD on Gumroad. Review checkout for tax or currency conversion.
Access: Agent-loaded on Agensi, not a standalone app. Downloadable access applies only where a separate download offer is linked. Package version 1.1.3.
Package requirements: Python 3.10+; Python 3.11+ for pyproject.toml. Retrieve the complete scanner and provider-rules reference. Version 1.1.2 supports the full entrypoint with rules supplied in memory in compatible no-save hosts; the host must permit Python execution. For the separate download, extract the complete folder. Missing or invalid rules return CANNOT_ASSESS.
You get: Local scanner, editable provider patterns, risky fixture, tests, and retirement-handoff instructions.
Licence and purchase details
Review the package licence and seller terms at checkout. An AI-agent subscription or setup service is not included.
Agensi access: In a creator test on October 17, 2026, Claude retrieved this package’s helper and required supporting files through Agensi and executed synthetic cases in its local Python sandbox. This was not Agensi-hosted execution, a complete package test, or a guarantee for other hosts. Local setup examples below apply to the separate downloadable package.
Try the included example first
For the separately obtained downloadable package, extract the entire ZIP, keep scripts and references together, and open a terminal in the extracted skill folder. On Windows, use py -3 instead of python if needed.
python scripts/scan_model_inventory.py fixtures/risky-sampleWhat you should see
The packaged risky fixture returns seven findings: two detected model references plus review flags for hardcoding, configuration indirection, inconsistent IDs and detected provider concentration. These are fixture results, not a complete inventory of a real deployment.
Then use your own evidence
Review the detected strings against actual deployment records before using the JSON handoff. Unknown provider or platform values must stay unknown until you have evidence. See references/retirement-handoff.md.
Reviewed local source and configuration. .env files, including examples, are excluded; other secret files are not guaranteed to be excluded. Supply only non-secret inputs.
Not runtime discovery, a complete bill of materials, or proof that a fallback is missing. Missing, linked or unreadable inputs must not be treated as a clean result.
// choose with context
Is Model Inventory Auditor the right skill?
It returns
File and line findings, detected model strings, and a JSON retirement handoff with unresolved deployment scope.
Do not use it as
Not runtime discovery, a complete bill of materials, or proof that a fallback is missing.
Compatibility, access, version, and licence
- Reads or accesses
- Reviewed local source and configuration. .env files, including examples, are excluded; other secret files are not guaranteed to be excluded. Supply only non-secret inputs.
- Compatibility
- Python 3.10+; Python 3.11+ for pyproject.toml. Retrieve the complete scanner and provider-rules reference. Version 1.1.2 supports the full entrypoint with rules supplied in memory in compatible no-save hosts; the host must permit Python execution. For the separate download, extract the complete folder. Missing or invalid rules return CANNOT_ASSESS.
- Version
- Agensi: package v1.1.3 (platform release v1.4). Gumroad download: v1.1.3. Checked October 23, 2026.
- Licence
- Review the package licence and seller terms at the linked destination.
Questions
Does a scan find every dependency?
No. Pattern matching can miss wrappers, indirect references and runtime configuration. Treat the output as incomplete, review deployment evidence, and never interpret an empty scan as proof of no dependencies.
Review the evidence before changing models.
Choose your access route above and start with a reviewed example.
Get the Model Inventory Auditor on Agensi