The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the 1password-pp-cli and 1password-pp-mcp tools from the vendor's GitHub repository (github.com/mvanhorn) and NPM package (@mvanhorn/printing-press-library). These tools are required for the skill's secret management and auditing capabilities.
[DATA_EXFILTRATION]: The 1password-pp-cli tool includes a --deliver webhook:<url> feature that allows sending command results, which may include secret metadata or resolved references, to a remote HTTP endpoint. Additionally, the feedback command can be configured via environment variables (API_1PASSWORD_FEEDBACK_ENDPOINT) to transmit data to a remote server.
[INDIRECT_PROMPT_INJECTION]: The skill provides an interface to process untrusted local data and natural language queries, which serves as a potential vector for indirect prompt injection.
Ingestion points: Data ingested via the --in-file, --query, and --task arguments in SKILL.md, as well as stdin for the feedback command.
Boundary markers: The skill emphasizes data redaction and reveal gates for secret values but does not provide specific prompt delimiters or instructions for the agent to ignore embedded commands in the source data during resolution or audit tasks.
Capability inventory: The skill has access to the local file system (via Read Bash), interacts with sensitive 1Password secret metadata, and supports network operations through the webhook delivery sink.
Sanitization: While the skill mentions policy checks and classification logic, it does not explicitly define sanitization routines to filter malicious prompt instructions within ingested content.

pp-1password