kj-aiml/heli-harness/heli-init

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Use when bootstrapping project facts/profile context for a Heli v0.11 linked project or an embedded compatibility workspace.

Install

npx vskill@latest install kj-aiml/heli-harness/heli-init

SKILL.md

Heli Init

Bootstrap or refresh descriptive project/repository facts without changing product source.

1. Resolve layout first

Run:

bash
heli status
  • Linked v0.11 project: project config/overlays live under .heli/; use Heli CLI state to resolve target/resource context.
  • Embedded compatibility workspace: use .heli-harness/ profile/workspace paths.

Do not infer linked authority from old embedded state files.

2. Inspect the project/repository

  • Read repo-local docs, package/build/test configuration, and relevant source structure.
  • Identify build/test/package-manager facts from evidence.
  • Do not edit product source code as part of profile initialization.
  • Preserve dirty user work.

3. Create or update the profile

Linked mode:

text
.heli/profiles/<repo>.md

Embedded compatibility:

text
.heli-harness/profiles/<repo>.md

Record:

  • observed stack;
  • existing patterns;
  • recommended conventions;
  • known tech debt;
  • forbidden patterns;
  • safer alternatives;
  • command tiers/risk notes;
  • repo risks;
  • exceptions;
  • evidence paths;
  • policy references.

Observed code is a fact, not automatically a recommendation.

4. Durable work record only when needed

Profile initialization does not require a named task merely to become authorized.

Create/update a durable work record when the work spans sessions, needs handoff/coordination, or carries significant verification/diagnosis obligations. In embedded compatibility mode, the existing task-state workflow remains available.

Command classification

Classify discovered commands by actual side effects/cost:

  • read/audit only;
  • non-mutating validation;
  • local mutation/build;
  • network/API/cost-bearing;
  • release/publish/deploy;
  • destructive/secret/outside-scope.

Risk tiers summarize impact; they do not grant authority.

Safety

  • Do not edit target source during profile bootstrap.
  • Do not commit/push/release.
  • Do not install dependencies or spend API credits without the applicable policy/grant.
  • Confirm target/resource identity when ambiguous.
  • Do not copy credentials or secret values into profiles/evidence.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/kj-aiml/heli-harness/heli-init/badge)](https://verified-skill.com/skills/kj-aiml/heli-harness/heli-init)

Something wrong with this skill or its page? Tell us.