bybren-llc/safe-agentic-workflow/templates

{{Title}}

This is a map-card. The skill's procedure lives in the skill definition cited below — do not restate its steps here. The card makes the skill discoverable from the graph; the definition is what actually runs.

Overview

{{3-5 sentences: what the skill does, its invocation trigger, and its output. State any precondition that must hold before it is safe to run.}}

Routes To

{{2-6 bullets: what this skill delegates to or produces — sub-skills, scripts, or the artifacts it writes. Name scripts and code files as plain inline code; never link to code files.}}

• {{PATH}} — {{what it does in this flow}}

Used By Roles

{{2-5 bullets: the agent roles that invoke this skill and at which point in their work. Link only to concept IDs listed in manifest.json; if a role concept does not exist yet, name it in prose without a link and report it as a suggestion.}}

• {{Role Title — when and why it invokes this}}

Citations

{{1-4 bullets. A stub type MUST carry an out-of-bundle citation to its source of truth — the validator enforces this. Out-of-bundle links appear ONLY in this section.}}

• {{Skill Definition}} — the authoritative procedure