The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from project-level configuration files and existing specification documents, which creates a surface for indirect prompt injection if those files are modified to include malicious instructions.
Ingestion points: Reads project-specific background, conventions, and constraints from openspec/config.yaml or config.yml. It also reads existing artifact files resolved via the openspec status command.
Boundary markers: The instructions do not define specific delimiters or instructions for the agent to ignore embedded commands within the content it reads.
Capability inventory: The skill is authorized to execute shell commands via the openspec CLI and has the capability to write design-related artifacts (proposals, designs, specs) to the local file system.
Sanitization: There is no evidence of validation or sanitization of the content retrieved from external files before it is processed by the agent.
[COMMAND_EXECUTION]: The skill invokes the openspec command-line tool to interact with the repository and manage state.
Evidence: Utilizes commands including openspec list, openspec store list, openspec status, openspec new change, and openspec instructions.

openspec-explore