The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides standard git and curl command templates for interacting with Bitbucket repositories and its REST API (e.g., git push, curl with authorization headers).
[PROMPT_INJECTION]: The instructions include behavioral directives wrapped in <IMPORTANT> tags (e.g., "ALWAYS use the Bitbucket API for operations instead of a web browser"). These are benign operational constraints intended to ensure the agent uses the provided tools correctly.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface due to its interaction with external repository data.
Ingestion points: Reads repository content and pull request data from Bitbucket (SKILL.md).
Boundary markers: No delimiters or instructions to ignore embedded content are used when processing external data.
Capability inventory: The skill can execute arbitrary git and curl commands and utilize the create_bitbucket_pr tool (SKILL.md).
Sanitization: No explicit sanitization or validation of ingested content is mentioned.
[CREDENTIALS_UNSAFE]: The skill references the BITBUCKET_TOKEN environment variable and suggests a troubleshooting step to embed this token into the git remote URL via git remote set-url. While this persists the sensitive token in the local .git/config file, it is a standard practice for authenticating git CLI operations when SSH is unavailable and the token itself is sourced from an environment variable rather than being hardcoded.

bitbucket