The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill instructs the user or agent to store GitCode personal access tokens in a plain-text configuration file at ~/.config/gitcode-cli/config.json. Additionally, the provided automation script in gitcode-pr-workflow.md recommends embedding the authentication token directly into the Git remote URL (e.g., https://{username}:{token}@gitcode.com/...). This practice can lead to credential leakage in shell history, logs, or the .git/config file, even though a warning is provided.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @oh-gc/cli package from the NPM registry. This is a third-party dependency necessary for the skill's functionality, but it introduces a supply chain risk as it is not an officially vetted or well-known service tool from the trusted list.
[COMMAND_EXECUTION]: The skill relies heavily on shell command execution to interact with the GitCode API and the local file system. The gitcode-pr-workflow.md file contains a complex shell script that performs automated Git operations, CLI configuration, and PR creation.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external GitCode repositories, such as issue descriptions, pull request comments, and commit messages.
Ingestion points: External data enters the agent context through commands like oh-gc issue view and oh-gc pr list (found in SKILL.md).
Boundary markers: The instructions do not specify any delimiters or safety warnings to ignore instructions embedded in the ingested repository content.
Capability inventory: The skill provides capabilities to write to the file system, modify repository settings, and perform network operations via the oh-gc CLI (found in SKILL.md).
Sanitization: There is no evidence of sanitization or validation of the ingested content before it is processed by the agent.

gitcode-cli