The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill accesses a local environment file at /Users/danielhowells/.codex/plugins/secrets/1password-service-account.env to retrieve a 1Password service account token. This token is used to non-interactively fetch Linear API keys via the op CLI, which is a standard practice for secure credential management in automation.
[COMMAND_EXECUTION]: The skill executes the op CLI tool using Python's subprocess.run and utilizes curl within a shell script to interact with the Linear GraphQL API. These executions are scoped to the skill's intended purpose of secret retrieval and API communication.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to https://api.linear.app/graphql, which is the official and well-known endpoint for the Linear service.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external data from API responses, creating a surface for indirect prompt injection. \n
Ingestion points: API responses from Linear are processed by scripts/check-response.py. \n
Boundary markers: Skill instructions explicitly command the agent to choose accounts based on read-only lookups and identify ambiguities before proceeding. \n
Capability inventory: Capabilities include file system access for local configuration and network access via curl. \n
Sanitization: The check-response.py script validates the JSON structure and checks for both GraphQL level errors and application level success flags in the data.

linear