Claude-skill-registry endpoint-exploration

Document REST API endpoints through systematic empirical testing. Analyzes URL structure, generates minimal test cases, executes requests, and produces concise API documentation. Use when exploring undocumented APIs, reverse-engineering endpoints, or creating integration documentation.

install
source · Clone the upstream repo
git clone https://github.com/majiayu000/claude-skill-registry
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/endpoint-exploration" ~/.claude/skills/majiayu000-claude-skill-registry-endpoint-exploration && rm -rf "$T"
manifest: skills/data/endpoint-exploration/SKILL.md
source content

Endpoint Exploration

Create comprehensive API endpoint documentation through systematic empirical testing.

Process

Copy this checklist and track your progress:

Endpoint Exploration Progress:
- [ ] Step 1: Analyze URL structure
- [ ] Step 2: Identify parameters and types
- [ ] Step 3: Generate minimal test cases
- [ ] Step 4: Execute tests and save samples
- [ ] Step 5: Analyze response patterns
- [ ] Step 6: Write documentation
- [ ] Step 7: Review for redundancy and clarity

1. URL Analysis

Parse the URL to identify path parameters (e.g.,

{id}
,
{format}
) and endpoint structure.

Do not make assumptions about what the API does - focus only on the URL structure.

2. Parameter Identification

For each parameter identified:

  • Determine likely type from URL structure
  • List normal values (valid inputs)
  • List invalid values (wrong types, edge cases)
  • Note if parameter appears optional vs required

3. Test Case Generation

Create a minimal test set covering:

  • Valid cases: 1-2 examples with real data
  • Format variations: If format is a parameter, test each option
  • Edge cases: Boundary values (0, 1, very large numbers, leading zeros)
  • Invalid cases: Wrong types, negative values, unsupported formats
  • Header tests: Accept header variations, conflicts
  • HTTP method tests: POST, PUT, DELETE (if GET endpoint)
  • Query parameter tests: Test if query params work

Avoid redundancy: If multiple values test the same behavior, pick one.

4. Sample Collection

Execute tests and save full HTTP responses (headers + body) to

samples/
directory:

  • Use
    curl -i
    to capture headers
  • Number files sequentially:
    01-description.ext
    ,
    02-description.ext
  • Use descriptive names indicating what each test validates

5. Response Analysis

For each unique response pattern, document:

  • HTTP status code
  • Content-Type header
  • Response structure
  • Any caching or rate-limit headers
  • Error message formats

Identify:

  • How to detect success vs failure
  • How to distinguish different error types
  • Any quirks (e.g., 200 for not-found, different error messages)

6. Documentation Writing

Create

endpoint-documentation.md
following the template structure with these sections:

Required sections:

  • Endpoint (HTTP method + full URL)
  • Authentication (if any)
  • Parameters (path, headers, query - if applicable)
  • Response Codes (table format)
  • Response Structure (tables preferred over examples)

Include only if discovered:

  • Caching behavior
  • Format selection rules
  • Special detection logic (e.g., how to tell if resource exists)
  • Important quirks or edge cases

Avoid:

  • Speculation or assumptions
  • Redundant information (each fact appears once)
  • Examples that just repeat table/prose information
  • Over-documentation of obvious behavior

7. Iterative Refinement

Review documentation against this checklist:

  • Redundancy: Each fact appears exactly once
  • Evidence: Only documented tested behavior
  • Clarity: Concise, no unnecessary detail
  • Completeness: All discoveries documented

Remove sections that don't add unique value.