using-apex-mcp_skill

This skill helps you leverage APEX MCP tools for pattern intelligence, task tracking, and reflections to improve code collaboration.
  • HTML

2

GitHub Stars

3

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill benredmond/apex --skill using-apex-mcp

  • apex-patterns-overview-guide.md18.0 KB
  • apex-reflect-guide.md15.6 KB
  • SKILL.md10.8 KB

Overview

This skill helps you call APEX MCP tools for pattern discovery, task tracking, and submitting reflections. It consolidates the behavior and intent of all 13 MCP tools, explains trust score meaning, and guides common workflows and error fixes. Use it to integrate pattern intelligence into developer workflows and to update pattern trust via reflections.

How this skill works

The skill maps each MCP tool to concrete operations: pattern lookup/discovery/explain, task lifecycle (create, update, checkpoint, complete), evidence appending, and reflection submission. It inspects task context, code context, and git evidence, returns trust-scored pattern packs, and validates reflections against strict rules (e.g., evidence format, outcomes). It also surfaces common errors and automatic preprocessor fixes.

When to use it

  • Start a new implementation or bug fix and need proven solutions (apex_patterns_lookup).
  • Explore semantic matches when simple lookup returns insufficient results (apex_patterns_discover).
  • Create and track a task from research through completion (apex_task_create → apex_task_update → apex_task_complete).
  • Submit formal outcomes and update pattern trust after committing changes (apex_reflect).
  • Get an intelligence pack with similar tasks, statistics, and recommended patterns (apex_task_context).

Best practices

  • Apply patterns with trust_score ≥ 0.7 confidently; validate patterns < 0.5 before applying.
  • Commit code and include git evidence (git_lines or commit) before calling apex_reflect.
  • Use batch_patterns for simple reflections; use claims only for advanced, structured submissions.
  • Track phase transitions with apex_task_update and frequent apex_task_checkpoint entries.
  • Cache lookup/discover responses (5-minute TTL) to avoid rate limits and speed workflows.

Example use cases

  • Implement JWT authentication: create a task, lookup builder-phase patterns, update files, complete and reflect with git evidence.
  • Fix a timeout bug: create a bug task, find similar tasks, append evidence referencing applied pattern, then submit reflection.
  • Add caching: use apex_patterns_discover to find high-trust cache invalidation patterns and apply with confidence.
  • Validate a new pattern: run apex_task_context to collect usage stats and then use apex_reflect with detailed claims to update trust.

FAQ

Prefer patterns rated ★★★★☆ and ★★★★★ (trust ≥ 0.7) for direct application. Treat 0.3–0.5 patterns cautiously and avoid ≤ 0.3 unless experimenting.

Why did apex_reflect fail validation?

Common causes: not committing changes, invalid outcome vocabulary, mixing batch and claims formats, or missing git SHA/lines. Follow the quick guide: commit, use allowed outcome strings, and submit only one format.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational