spec-sync

Detects weegloo-server spec changes and syncs them into plugin skills and rules.

1|2|Updated Feb 20, 2026
One-click install
npx skills add https://github.com/weeglooapi/weegloo-mcp-plugin --skill spec-sync-weeglooapi
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: spec-sync
Source: https://github.com/weeglooapi/weegloo-mcp-plugin/tree/main/.claude/skills/spec-sync
Command: npx skills add https://github.com/weeglooapi/weegloo-mcp-plugin --skill spec-sync-weeglooapi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When the weegloo-server API changes (endpoints, MCP tools, error codes, script statement types), the agent-facing skills and rules in this repository silently drift out of date. Manually tracking what changed across a git range and deciding which documentation files to update is error-prone and tedious. ## Core Features & Use Cases - Mechanical change detection: Diffs operationIds, MCP tool groups, error codes, and schema-internal enums (e.g., Script statement types) between two git refs of the server repo, with dynamic module discovery instead of hardcoded paths. - Deployment gating: Cross-checks detected changes against dev-stack Swagger URLs per plane so only actually deployed features get documented, with MCP exposure triage for hidden [-Tag] tools. - Approval-gated editing: Produces a structured impact report (detected changes, affected files with rationale, items needing human judgment) and edits skill/rule files only after explicit approval. - Use Case: After a server release, run the skill with the server repo path, a git range, and Swagger URLs to get a reviewed list of exactly which skill and rule files need updating and why. ## Quick Start Ask the agent to sync the latest weegloo-server spec changes into the plugin skills, providing the server repo path, the base and head git refs, and the dev Swagger URLs.

Frequently Asked Questions about spec-sync

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I sync server API changes into agent skill documentation?

Provide the server repo path, a git range (BASE..HEAD), and dev-stack Swagger URLs. The skill mechanically extracts changed operationIds, MCP tool groups, error codes, and schema enums, then reports affected skill and rule files for your approval before editing.

How does the skill decide which server changes to document?

It applies a triage step asking whether each change alters agent behavior. Internal refactors, infrastructure, logging, and admin-only changes are excluded, while endpoints, MCP tools, validation rules, and agent-triggerable error codes are included.

Does it document features that are merged but not yet deployed?

No. Changes are gated against dev-stack Swagger definitions, which are the source of truth. Undeployed changes are reported as not deployed and only documented if a human explicitly instructs it, with the pre-deployment status disclosed.

Why does the skill avoid hardcoding module paths when scanning the server repo?

Modules move between packages, and a hardcoded list once caused nine live endpoints to be misreported as deleted. The skill dynamically discovers modules containing @Operation annotations at both git refs and unions the results.

Can the skill edit files before the impact report is approved?

No. Editing is strictly gated behind human approval of the impact report. Before input confirmation it may only run a small whitelist of read-only git commands, and it never touches files until the report is signed off.