The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The reference documentation provides code templates that introduce significant vulnerability surfaces due to missing sanitization of untrusted data.
Ingestion points: The Express example in references/express-middleware-integration.md ingests data from req.params.id. The Python example in references/python-sentry-integration.md ingests order_id from the route.
Capability inventory: Both examples perform database operations using these inputs.
Sanitization: Both examples fail to use parameterized queries, instead using template literals or f-strings to build SQL queries.
Evidence: The JavaScript example uses db.query(SELECT * FROM users WHERE id = ${userId}) and the Python example uses db.query(f'SELECT * FROM orders WHERE id = {order_id}'). These patterns are highly susceptible to SQL injection.
[COMMAND_EXECUTION]: The skill provides instructions and scripts for executing local commands and installing global packages.
SKILL.md and references/sentry-setup.md recommend npm install -g @sentry/cli and sentry init -d for environment setup.
references/source-maps-and-release-management.md includes a bash script that performs release management using the sentry-cli tool. These are standard operations for the tool's intended use.

error-tracking