{{SKILL_NAME}}

Invocation (template) — replace {{LAUNCHER}} with the real launcher from the Delegation Matrix. Do not paste workflow literals unless this skill is vc-workflow.

Path	Literal
1. Worker	vibecrafted {{LAUNCHER}} <agent>
2. Interactive	/vc-{{LAUNCHER}} — in-session; native when required
3. Operator	same worker form via vc-dispatch / operator lines

{{SKILL_NAME}} — TODO one-line tagline

Scaffolded {{CREATED_DATE}} via tools/vc-skill-new.sh. Replace every TODO marker before opening a PR.

Operator Entry

Living Tree / Worktree Rule

This workflow runs in the operator's current checkout and current branch. Do not create, switch to, or move execution into a git worktree unless the operator explicitly asks for one in this prompt. Re-read files before editing, adapt to concurrent changes, and report substrate failure if the tree is too poisoned to continue safely. The one sanctioned second mode is a Fleet Worktree dispatch (written plan, pre-committed verifiers, disjoint domains, single-thread integrator — see Living Tree Rule, Mode B); outside that formation, stay in the shared tree.

See Living Tree Rule.

Repository Work Doctrine

For repository work, start with Loctree as the map: use loct context, loct occurrences, loct body, and loct find --literal before broad manual search. Use AICX for intent and session context. Use rg/grep as fallback or local magnifier, not as a replacement for structural mapping. If Loctree fails or misses a surface, append feedback to ~/.vibecrafted/loctree/loctree-fail.md.

Standard launcher:

vibecrafted {{SKILL_NAME_NO_PREFIX}} claude --prompt 'TODO concrete operator example'
vc-{{SKILL_NAME_NO_PREFIX}} codex --prompt 'TODO shell-shortcut example'

Purpose

TODO — Replace this section. State the one outcome this skill produces. Skills exist to compress a recurring operator move into a named, repeatable surface. If this section reads like a list of capabilities, narrow it.

The bar from CONTRIBUTING-SKILLS.md: one sharp axis, not a Swiss-army knife.

When To Use

Trigger conditions (replace all bullets):

• TODO — primary operator situation where this skill is the right call
• TODO — secondary situation, if any
• TODO — explicit non-overlap with existing vc-* skills

When NOT to use:

• TODO — adjacent skill that handles a similar-but-distinct situation
• TODO — situation that should escalate to vc-implement or vc-marbles instead

Pipeline Position

Where does this fit in the Vetcoders workflow chain?

• Upstream: TODO (e.g. follows vc-init, runs after vc-research)
• Downstream: TODO (e.g. emits handoff for vc-release or vc-dou)

Acceptance Criteria

The skill run is done when:

• TODO — concrete, falsifiable check #1
• TODO — concrete, falsifiable check #2
• TODO — operator-visible deliverable (file, report, commit)

If any acceptance bullet cannot be ticked with evidence, the skill has not completed — say so explicitly in the final report.

Anti-Patterns

• TODO — common failure mode #1 (e.g. running this skill before vc-init)
• TODO — common failure mode #2 (e.g. expanding scope beyond the one sharp axis)
• Pasting workflow / ERi rails when this skill is not vc-workflow
• Inventing a fake vibecrafted <name> <agent> worker for a foundation skill
• Skipping the Living Tree re-read before edit when concurrent agents are active
• Claiming "done" without ticking the acceptance criteria above

Examples

See examples/example-prompt.md for a minimal trigger phrase + expected behavior pair.

Verify before the handoff

Before claiming "done", walk around the truck — see Verification Rule: real artifact, not green gates only; never trust upstream verification as proof; check your own instrument.

Progressive disclosure: keep this SKILL.md lean; put long procedures in references/ and load them only when the run needs them (Delegation Matrix, CONTRIBUTING-SKILLS).

𝚅𝚒𝚋𝚎𝚌𝚛𝚊𝚏𝚝𝚎𝚍. with AI Agents by Vetcoders (c)2024-2026 LibraxisAI