Claude Cursor Skill

feature-radar-validate

Validate SKILL.md frontmatter and .feature-radar/ files against format rules. Runs validate.sh, reports errors/warnings, and auto-fixes issues. MUST use this skill after editing any SKILL.md or .feature-radar/ file, even if the user doesn't ask — catches format bugs like the Agen

LLM Mart · 0 points · 0 views 0 listing impressions 0 install-command copies
Download runkids-my-skills-feature-radar_feature-radar-validate-614b89f.zip · 3 KB

Install

skills CLI npx skills add https://github.com/runkids/my-skills/tree/main/feature-radar/feature-radar-validate
Claude Code claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install runkids-my-skills@llmmart
Git git clone https://github.com/runkids/my-skills.git

The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole runkids/my-skills collection as a plugin from our marketplace. Git is the plain clone.

Skill manifest

Validate Feature Radar

Run skills/feature-radar-validate/scripts/validate.sh from the project root to check SKILL.md frontmatter and .feature-radar/ SPEC compliance, then fix any issues found.

Why This Matters

The description field in SKILL.md has a 1024-character limit in the Agent Skills spec (agentskills.io/specification, checked 2026-09-29). Agents that enforce the spec may reject a longer description; Claude Code instead truncates the skill listing at 1,536 characters, cutting off trigger text. Similarly, .feature-radar/ files must follow SPEC.md naming and metadata conventions or downstream tools can't parse them. This skill catches these issues before they cause problems.

Workflow

Step 1: Run Validation

bash skills/feature-radar-validate/scripts/validate.sh

Read the full output. Note the exit code:

  • Exit 0: all checks passed (may still have warnings)
  • Exit 1: errors found — must be fixed

Step 2: Report Results

Present the results clearly:

── Feature Radar: Validate ──

Errors:   {n}
Warnings: {n}

{List each error/warning with file path and issue}

If everything passes, say so and stop. No further action needed.

Step 3: Auto-Fix (if errors or warnings found)

For each issue, apply the appropriate fix:

Issue Fix Strategy
description > 1024 chars Trim to fit — cut the least essential trigger phrases or examples first, keep the core "what it does" and "Use when" intact. Show before/after char count.
description missing "Use when" Add a "Use when:" section based on the skill's purpose
name not kebab-case Rename to kebab-case
name missing Derive from directory name
name > 64 chars Propose a shorter name and ask user to confirm — the name is how the skill is invoked
name does not match directory Set name to the directory name; if the name is the intended one, ask user before renaming the directory (other skills reference it by path)
Body > 500 lines Flag for user — this requires judgment about what to extract into reference files
Filename not {nn}-{slug}.md Rename file to correct format
Missing **Status**: / **Impact**: / **Effort**: Add field with a placeholder value, ask user to confirm
base.md count mismatch Update the Tracking Summary table counts

After fixing, re-run bash skills/feature-radar-validate/scripts/validate.sh to confirm all errors are resolved.

Step 4: Completion Summary

── Feature Radar: Validate Complete ──

Files fixed:   ~ {path} ({what changed})
Errors fixed:  {n}
Warnings fixed: {n}
Remaining:     {n} (need user input)

Proactive Triggering

When you notice yourself editing skills/*/SKILL.md or .feature-radar/**/*.md, run validation afterward without being asked. A quick bash skills/feature-radar-validate/scripts/validate.sh check takes seconds and prevents silent breakage.

Files (my-skills)
  • scripts
    • validate.sh 5.4 KB
      Loading…
  • SKILL.md 3.5 KB
    ---
    description: |-
      Validate SKILL.md frontmatter and .feature-radar/ files against format rules. Runs
      validate.sh, reports errors/warnings, and auto-fixes issues. MUST use this skill after
      editing any SKILL.md or .feature-radar/ file, even if the user doesn't ask — catches format
      bugs like the Agent Skills spec's 1024-char description limit before they break the skill.
      Use when the user asks to validate, lint, or verify skills or .feature-radar/ files, after
      editing either, or before committing changes that touch skills/ or .feature-radar/.
    metadata:
      targets:
        - claude
    name: feature-radar-validate
    ---
    
    # Validate Feature Radar
    
    Run `skills/feature-radar-validate/scripts/validate.sh` from the project root to check SKILL.md frontmatter and `.feature-radar/` SPEC compliance, then fix any issues found.
    
    ## Why This Matters
    
    The `description` field in SKILL.md has a 1024-character limit in the Agent Skills spec (agentskills.io/specification, checked 2026-09-29). Agents that enforce the spec may reject a longer description; Claude Code instead truncates the skill listing at 1,536 characters, cutting off trigger text. Similarly, `.feature-radar/` files must follow SPEC.md naming and metadata conventions or downstream tools can't parse them. This skill catches these issues before they cause problems.
    
    ## Workflow
    
    ### Step 1: Run Validation
    
    ```bash
    bash skills/feature-radar-validate/scripts/validate.sh
    ```
    
    Read the full output. Note the exit code:
    - **Exit 0**: all checks passed (may still have warnings)
    - **Exit 1**: errors found — must be fixed
    
    ### Step 2: Report Results
    
    Present the results clearly:
    
    ```
    ── Feature Radar: Validate ──
    
    Errors:   {n}
    Warnings: {n}
    
    {List each error/warning with file path and issue}
    ```
    
    If everything passes, say so and stop. No further action needed.
    
    ### Step 3: Auto-Fix (if errors or warnings found)
    
    For each issue, apply the appropriate fix:
    
    | Issue | Fix Strategy |
    |-------|-------------|
    | `description` > 1024 chars | Trim to fit — cut the least essential trigger phrases or examples first, keep the core "what it does" and "Use when" intact. Show before/after char count. |
    | `description` missing "Use when" | Add a "Use when:" section based on the skill's purpose |
    | `name` not kebab-case | Rename to kebab-case |
    | `name` missing | Derive from directory name |
    | `name` > 64 chars | Propose a shorter name and ask user to confirm — the name is how the skill is invoked |
    | `name` does not match directory | Set `name` to the directory name; if the `name` is the intended one, ask user before renaming the directory (other skills reference it by path) |
    | Body > 500 lines | Flag for user — this requires judgment about what to extract into reference files |
    | Filename not `{nn}-{slug}.md` | Rename file to correct format |
    | Missing `**Status**:` / `**Impact**:` / `**Effort**:` | Add field with a placeholder value, ask user to confirm |
    | `base.md` count mismatch | Update the Tracking Summary table counts |
    
    After fixing, re-run `bash skills/feature-radar-validate/scripts/validate.sh` to confirm all errors are resolved.
    
    ### Step 4: Completion Summary
    
    ```
    ── Feature Radar: Validate Complete ──
    
    Files fixed:   ~ {path} ({what changed})
    Errors fixed:  {n}
    Warnings fixed: {n}
    Remaining:     {n} (need user input)
    ```
    
    ## Proactive Triggering
    
    When you notice yourself editing `skills/*/SKILL.md` or `.feature-radar/**/*.md`, run validation afterward without being asked. A quick `bash skills/feature-radar-validate/scripts/validate.sh` check takes seconds and prevents silent breakage.
    

Comments (0)

Sign in to join the conversation.

No comments yet.

Reviews (0)

No reviews yet.