add-skill
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill workflow requires the execution of
make checkin Step 5 to validate newly created skills. This grants the agent the ability to run shell commands that depend on the state of the local environment and the content of the generated files.- [DATA_EXFILTRATION]: In Step 2, the skill is instructed to recursively read the.local/directory to gather source material. This presents a risk of sensitive data exposure if credentials, local configurations, or other private information are stored within that directory.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core functionality of ingesting external data to influence its output. - Ingestion points: The agent reads all files in the
.local/folder and performs web searches to research facts and documentation (Step 2). - Boundary markers: There are no instructions provided to use delimiters or ignore embedded instructions within the gathered research material.
- Capability inventory: The agent has the capability to write and modify files (Step 4, Step 7) and execute shell commands through the validation process (Step 5).
- Sanitization: No sanitization or verification steps are defined for the content retrieved from the web or local scratch files before it is processed or used in the creation of new skill instructions.
Audit Metadata