The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by dynamically inferring its verification workflow from external, potentially untrusted project data.
Ingestion points: Data is ingested from wrangler.toml, git diff, and git log (File: SKILL.md).
Boundary markers: No boundary markers or specific instructions are provided to isolate or ignore instructions that might be embedded in commit messages or configuration files.
Capability inventory: The agent possesses capabilities to execute network requests (curl), trigger deployments (wrangler), and run local tests (npm/bun).
Sanitization: There is no evidence of sanitization or validation for content extracted from the git history before it influences the agent's decision-making process.
[COMMAND_EXECUTION]: The skill executes various shell commands to manage the deployment lifecycle.
Evidence: The skill invokes wrangler deploy, wrangler tail, and wrangler secret list. These are standard tools for the Cloudflare Workers ecosystem and are used here for their intended purpose.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to verify the state of the deployment.
Evidence: The curl command is used to hit endpoints on dynamically generated worker URLs to verify responses and headers.

deploy-verify