meta-config-stack-detect
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides the user with a pre-formatted CLI command (
npx agents-inc share --stdin) to publish the detected configuration. This ensures that the final action of sharing data is explicitly initiated and controlled by the user. - [EXTERNAL_DOWNLOADS]: The skill describes fetching a library catalog from external marketplace repositories or local caches. These are standard functional resources used to map detected project libraries to known internal skill identifiers. The skill author ('agents-inc') manages these repositories.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from untrusted project files such as manifests (
package.json), lockfiles (pnpm-lock.yaml), and framework configs (vite.config.ts). - Ingestion points: Manifest files, lockfiles, and framework-specific configuration files across workspaces.
- Boundary markers: Not explicitly defined for file ingestion; however, the logic relies on structured pattern matching against a vendor-supplied catalog rather than open-ended instruction following.
- Capability inventory: The skill itself has no direct network, file-writing, or execution capabilities; it only emits text for the user to review.
- Sanitization: The resulting configuration is validated against a strict schema by the vendor's CLI tool, which discards unauthorized or unknown keys.
Audit Metadata