- Home
- Skills
- Technical Documentation
- Sync wiki page
Sync wiki page
Maintains a structured, session-aware project wiki with automated indexing and template enforcement.
Free
Works with the AI Agents you already use
Sync wiki page
Example session with this skill installed
Run wiki sync to document the fix for the API 403 errors we found in the auth middleware during this session.
- Read your context and instructions
- Compiled the sync wiki page
I have updated the wiki page for this session to document the authentication middleware fixes.
2026-05-12-fix-api-auth-403.md
---
session_id: ses_a9b8c7d6
date: 2026-05-12
type: debug
services: [api, auth-service]
tickets: [DEV-402]
tags: ['403', middleware, security]
branch: "fix/auth-headers"
related: []
---
## TL;DR
Resolved a race condition in the JWT validation middleware that caused intermittent 403 Forbidden errors when processing rapid concurrent requests.
## Investigation
The middleware was failing to await the public key retrieval from the cache, leading to unverified tokens during high-traffic spikes.
## Fixes
- Added `await` to the `getPublicKey` call in `src/middleware/auth.js`.
- Implemented a retry mechanism for transient cache misses.
## Test Results
- **Unit Tests** — Passed (12/12)
- **Integration Tests** — Passed with 100 concurrent requests.
## Update — 2026-05-12 14:45
Verified the fix in the staging environment. No 403 errors observed in the last 30 minutes of logs.
## Follow-ups
- Monitor error rates in production after deployment.
- Refactor cache TTL settings for the public key store.
Next steps
- Review the updated
wiki/INDEX.mdto ensure the summary is accurate. - Commit the changes to the
wiki/directory.
Connects securely to your tools. The creator never sees your data.
About this skill
This skill manages technical documentation by centralizing session reports into a structured wiki. It eliminates the manual overhead of tracking changes across multiple interactions by enforcing a single-file-per-session rule. The agent handles the entire lifecycle of a wiki entry: scanning for existing session context, generating unique filenames, applying standardized templates, and maintaining a global index.
The system avoids the common pitfalls of AI-generated documentation by strictly forbidding Markdown tables in favor of git-diff-friendly bullet lists and cards. It ensures data integrity through a triple-source resolution process, checking conversation history, YAML frontmatter, and a dedicated session mapping file before writing any data. This redundancy prevents accidental overwrites and maintains a clean history of investigations, implementations, and code reviews.
What it does
- Session tracking ensures all updates within a single session write to the same file via a unique session ID.
- Automated scaffolding creates a
wiki/pages/directory and applies structural templates for debug, investigation, and code-review types. - Index management maintains a chronological
INDEX.mdwith links and summaries, automatically inserting new entries at the top. - Conflict resolution performs a triple-check of history, frontmatter, and JSON mappings to prevent duplicate or conflicting page writes.
- JSON mapping manages
wiki/.sessions.jsonto link session IDs to physical files for persistent cross-tool reference.
How it works
- Resolve Session: The agent scans conversation logs, YAML frontmatter in existing pages, and the
.sessions.jsonmap to identify the correct target file. - Scaffold or Load: If no page exists, it generates a dated kebab-case filename; otherwise, it loads the existing session file.
- Apply Template: Content is formatted based on
wiki/TEMPLATE.mdpresets, ensuring mandatory fields like TL;DR and branch info are included. - Update Index: The
wiki/INDEX.mdand.sessions.jsonfiles are refreshed to reflect the new or modified content.
Frameworks & tools
This skill operates on standard filesystem structures using Markdown (GFM), YAML frontmatter, and JSON. It relies on a local wiki/ directory structure containing pages/, TEMPLATE.md, and INDEX.md.
Why this beats prompting it yourself
Prompting an LLM to "update the wiki" often leads to fragmented files or inconsistent formatting that breaks your site's searchability. This skill enforces a rigid resolution logic and a specific data schema that ensures your documentation remains a reliable source of truth across weeks of development.
Use cases
- Documenting a complex bug investigation across multiple debug attempts.
- Recording code review findings and implementation steps for a specific feature branch.
- Maintaining a chronological log of architectural decisions within a project wiki.
Known limitations
Does not support Markdown tables to ensure clean git diffs. Requires a pre-existing wiki/TEMPLATE.md and wiki/README.md to function correctly within the scaffold.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Technical Documentation.
No support for GFM tables. Requires wiki/TEMPLATE.md and wiki/README.md to exist. Limited to one wiki page per session ID.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 20 days ago
- Free with an Agensi account