The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill is designed to automatically harvest sensitive credentials by executing doppler secrets get AWS_ACCESS_KEY_ID --plain and reading AWS environment variables.
[DYNAMIC_CONTEXT_INJECTION]: The skill uses the ! syntax in SKILL.md to execute the binary ${CLAUDE_PLUGIN_ROOT}/bin/ops-external immediately upon loading the skill. This allows for code execution before the user or agent can review the logic.
[COMMAND_EXECUTION]: The skill makes extensive use of the Bash tool to run the AWS CLI (aws ce), curl for Shopify API calls, and local file operations via cat and jq.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill fetches and processes data from external APIs (AWS Cost Explorer, Shopify Admin API) and a local registry.json file.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore potentially malicious content embedded in the external API responses.
Capability inventory: The skill has access to the Bash, Write, and WebFetch tools, creating a chain where external data could potentially trigger command execution or file writes.
Sanitization: No sanitization or validation logic is applied to the data retrieved from Shopify or AWS before it is incorporated into the dashboard context.

ops-revenue