The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The scripts/ingest.py script fetches UI component documentation and deep-dive analysis files from a public GitHub repository (github.com/inbn/component-gallery) and crawls component.gallery using an external service. These downloads target well-known and appropriate sources for the skill's stated purpose.
[COMMAND_EXECUTION]: The skill uses subprocess.run in scripts/ingest.py and scripts/query.py to execute external CLI tools including firecrawl (for web crawling), rlama (for retrieval-augmented generation tasks), and a local retrieval script. These operations are core to the skill's functionality for building and querying a local search index.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests and processes untrusted data from the web.
Ingestion points: scripts/ingest.py (crawls external web pages and fetches markdown files from GitHub).
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the indexing logic.
Capability inventory: The skill possesses command execution capabilities through subprocess.run calls in scripts/ingest.py and scripts/query.py.
Sanitization: No validation or sanitization of the downloaded content is performed before it is indexed and subsequently presented to the agent for synthesis.

component-gallery