specsfy-02-backlog

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Node.js script (scripts/iniciar_backlog.mjs) to automate the creation of backlog files within the project's specs/backlog/ directory.\n- [EXTERNAL_DOWNLOADS]: The documentation mentions installing optional specialist skills using npx skills add from the vendor's GitHub repository (github.com/promovaweb/specsfy). This is a standard mechanism for extending the framework's capabilities using the author's own resources.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from various project files (such as specs/inbox/ and docs/) and user input to generate backlog briefs. This constitutes an attack surface where instructions could theoretically be embedded in the source documents.\n
  • Ingestion points: The agent reads from specs/inbox/*.md, specs/backlog/*.md, docs/**/*.md, and MVP.md.\n
  • Boundary markers: There are no explicit instructions to the agent to treat the content of these files as untrusted or to use specific delimiters.\n
  • Capability inventory: The skill has the ability to write files to the local filesystem via the iniciar_backlog.mjs script.\n
  • Sanitization: The script performs kebab-case validation on slugs used for filenames, which prevents path traversal attempts.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 10:09 PM
Security Audit — agent-trust-hub — specsfy-02-backlog