The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions require the agent to execute shell commands using curl to interact with the ServiceGraph REST API endpoints.
[DATA_EXPOSURE]: The skill specifies accessing a local sensitive file, .env.local, via shell commands to load API credentials. While the skill includes instructions to avoid reading these secrets directly into the LLM context to protect the token, the execution environment still performs reads on a sensitive local path.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from an external API (api.servicegraph.co), which presents a vulnerability surface if the retrieved data (such as agency names or descriptions) contains malicious instructions.
Ingestion points: Data is retrieved from the search, unlocks, and :apex endpoints in the pro_services dataset.
Boundary markers: The instructions do not define specific delimiters or warnings to ignore embedded instructions in the API responses.
Capability inventory: The agent has capabilities for shell execution (via curl), network access to api.servicegraph.co, and reading the local .env.local file.
Sanitization: There are no instructions for sanitizing or validating external content before it is processed by the agent.

find-pr-agency