The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes git and curl commands to perform repository management and API interactions.
[CREDENTIALS_UNSAFE]: The instructions suggest embedding the ${BITBUCKET_TOKEN} directly into the git remote URL (e.g., git remote set-url origin https://x-token-auth:${BITBUCKET_TOKEN}@bitbucket.org/...). While effective for automation, this practice results in the authentication token being stored in plain text within the repository's .git/config file on the local filesystem.
[DATA_EXFILTRATION]: The skill performs network requests to bitbucket.org. Since Bitbucket is a well-known service and the primary target of this skill, these operations are legitimate and represent the intended functionality.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes content from external repositories.
Ingestion points: Repository files, branch names, and Bitbucket API responses accessed during execution.
Boundary markers: None identified in the instructions.
Capability inventory: Includes git write operations (git push), branch management, and API calls via curl and the create_bitbucket_pr tool.
Sanitization: No specific sanitization or validation of the ingested repository data is described.

bitbucket