capture — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited capture (Agent Skill) and scored it 96/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 1 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 1 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
You are capturing a product feature into Sprintra project management. Your job is to have a SHORT conversation (2-3 exchanges max) to understand what the user wants, then create well-structured artifacts via MCP tools. Do NOT over-interview — developers hate long Q&A. Get enough to create quality artifacts, infer the rest.
The user described: $ARGUMENTS
If the description is clear enough to act on, skip to Step 2. Only ask if genuinely ambiguous. When you do ask, batch all questions into ONE message:
If the user gave enough context already, say "Got it, let me capture this" and proceed.
Call the features MCP tool:
features(method: "create", title: "...", description: "...", priority: "...", acceptance_criteria: [...])Rules:
inferred_fields arrayFor each distinct piece of work, call:
stories(method: "create", feature_id: "...", title: "...", type: "story|task|bug", story_points: N, description: "...", acceptance_criteria: [...])Rules:
story (user-facing), task (technical), bug (fix), chore (maintenance)If the conversation included technical choices (e.g., "let's use PostgreSQL", "we'll go with REST not GraphQL"), record each as:
decisions(method: "add", title: "...", context: "...", decision: "...", consequences: "...", category: "architecture|technology|design|infrastructure|process")Show a compact summary:
Feature: [title] ([priority])
Stories: [count] ([total points] SP)
1. [story title] — [points] SP
2. [story title] — [points] SP
...
Decisions: [count] recorded
Completeness: [score]%/capture with no args, ask "What are we building?" — one sentence is enough.~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.