The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill's documentation provides installation instructions using npx and git clone from the vendor's official GitHub organization (code-peekr). These are standard distribution methods for the tool and target the vendor's own infrastructure.
[COMMAND_EXECUTION]: The skill suggests using git remote get-url origin to automatically infer the repository name for tool calls. This is a read-only operation intended to provide context to the agent and does not pose a security risk in this context.
[DATA_EXFILTRATION]: The skill core functionality involves sending queries to a hosted API (codepeekr.dev). While this involves sending codebase-related queries to an external service, it is the stated primary purpose of the skill and requires explicit user configuration of an MCP server and authentication token.
[PROMPT_INJECTION]: The skill processes research results returned by the ask_repo tool. There is a risk of indirect prompt injection if the codebase being analyzed contains malicious instructions designed to influence the agent's summary of the research.
Ingestion points: SKILL.md (instructions for processing output from mcp__peek__ask_repo)
Boundary markers: Absent in the prompt templates.
Capability inventory: The agent retains its full suite of local file-editing and command execution tools while using this skill.
Sanitization: The skill does not specify sanitization for the external service's output.

peek