agentleFS
477174477174/interface-mcp-agent-creator/.claude/skills/feature-mapper/SKILL.md

feature-mapper

Skill0 starsChanged 8 months agoMITView on GitHub

Transform app exploration data into structured feature documentation organized by user role. Creates a feature matrix showing which roles can access which features and what actions each role can perform. Use when asked to: document features, map features by role, create feature matrix, what can each role do, compare user roles, feature documentation.

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. About 0 are common, 11 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈11

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Check for exploration artifacts at .claude/artifacts/exploration-*/.

    line 28
  2. Distinctive:

    If found, read available exploration files and proceed with extraction.

    line 29
  3. Distinctive:

    If not found, request one of:

    line 30
  4. Distinctive:

    Treat each page URL as a potential feature area.

    line 46
  5. Very specific:

    When multi-role data is present, apply references/role-diff-algorithm.md:

    line 84
  6. Distinctive:

    If only one role is available, skip pairwise role comparison and mark all extracted features against that role only.

    line 98
  7. Distinctive:

    If user updates are provided, revise feature boundaries, categories, actions, or role mappings.

    line 104
  8. Very specific:

    Follow references/output-format.md schema exactly.

    line 111
  9. Distinctive:

    Ensure IDs are kebab-case and stable because downstream tool generation depends on them.

    line 112
  10. Distinctive:

    No exploration data found: ask user to run app-explorer or provide manual feature descriptions.

    line 116
  11. Distinctive:

    Do not perform accessibility auditing.

    line 122
  12. Distinctive:

    Do not generate test cases.

    line 123
  13. Distinctive:

    Do not re-explore the application.

    line 124
  14. Distinctive:

    Do not invoke browser tools from this skill.

    line 125

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Feature Mapper

Overview

This skill transforms app exploration output into structured feature documentation with role-aware access details. It produces a feature matrix that maps feature availability and actions by role, then saves the result in a standardized artifact format. It can run from app-explorer artifacts or from manual feature descriptions provided by the user.

References

  • references/chrome-mcp-compat.md
  • references/output-format.md
  • references/feature-categorization.md
  • references/role-diff-algorithm.md

Prerequisites

  • Input source is one of:
    • App exploration output under .claude/artifacts/exploration-*/
    • Manual feature descriptions supplied by the user
  • This skill processes data only. It does not require browser automation and does not run Chrome MCP tools.

Input Detection

  1. Check for exploration artifacts at .claude/artifacts/exploration-*/.
  2. If found, read available exploration files and proceed with extraction.
  3. If not found, request one of:
    • User runs app-explorer first, or
    • User provides manual feature descriptions
  4. Accept both single-role and multi-role input:
    • Single-role: build features and role access for one role
    • Multi-role: build features plus role comparison output

Feature Extraction Workflow

Step 1: Read Input

  • Parse exploration data fields such as pages, elements, and navigation_edges.
  • For manual input, normalize into the same internal shape: page context, UI elements, and role labels when available.

Step 2: Group by Page

  • Treat each page URL as a potential feature area.
  • Merge closely related routes when they belong to one feature flow.

Step 3: Identify Features

  • Group related UI elements into logical features.
  • Example: user table + search input + create button + delete buttons -> User Management with sub-capabilities.

Step 4: Categorize

  • Assign one primary category per feature using references/feature-categorization.md.
  • Allowed categories: crud, form, dashboard, settings, search, auth, media, messaging, reporting, navigation, other.

Step 5: Extract Actions

  • Derive action names from controls and semantics.
  • Typical mappings:
    • Create button -> create
    • Search input -> search
    • Table/list -> view
    • Delete control -> delete
    • Save/edit flow -> edit

Step 6: Build Feature Entries

For each feature, produce:

  • id: kebab-case stable identifier
  • name: human-readable feature title
  • description: concise purpose
  • url: primary entry URL
  • category: one allowed value
  • actions: deduplicated action array
  • ui_elements: mapped references to source elements
  • roles: role-specific accessibility and available actions

Role-Based Analysis

When multi-role data is present, apply references/role-diff-algorithm.md:

  1. Compare page inventories per role to identify exclusive and shared pages.
  2. Compare element inventories for shared pages to detect role-only controls.
  3. Compare actions per shared feature to identify permission differences.
  4. Populate each feature's roles object:
    • accessible (boolean)
    • available_actions (array)
    • restricted_elements (array, optional)
  5. Build role_comparison entries containing:
    • exclusive features per role
    • shared features
    • action differences

If only one role is available, skip pairwise role comparison and mark all extracted features against that role only.

Iterative Refinement Loop

  1. Present the initial feature map draft.
  2. Ask: Does this look correct? Should I add, modify, or remove any features?
  3. If user updates are provided, revise feature boundaries, categories, actions, or role mappings.
  4. Re-present until approved.
  5. Save final output once approved.

Output

  • Save output to .claude/artifacts/feature-map-{timestamp}/.
  • Follow references/output-format.md schema exactly.
  • Ensure IDs are kebab-case and stable because downstream tool generation depends on them.

Troubleshooting

  • No exploration data found: ask user to run app-explorer or provide manual feature descriptions.
  • Unclear feature boundaries: fall back to URL-based grouping (each unique page path as one feature area).
  • Single role only: skip role comparison; keep role access populated for the available role.

Scope Guardrails

  • Do not perform accessibility auditing.
  • Do not generate test cases.
  • Do not re-explore the application.
  • Do not invoke browser tools from this skill.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.