kevmoo/kevmoo_skills · Archived

sidequest

>- Synthesizes conversation history and active tasks into a visual hierarchy map (`sidequest.md`) backed by a deterministic JSON state file (`sidequest.json`). Supports multiple sequential and concurrent main quests, sub-quests, and side-quests with automatic hierarchical numbering and completion sequencing. Use when the user invokes `/sidequest`, asks where we are, what we were doing, or what's on our stack, or when the conversation branches across multiple topics, blockers, or digressions. Do…

First seen Jul 12, 2026

Installation

$ npx skills add kevmoo/kevmoo_skills --skill sidequest

Stronger alternatives

This repository is archived — consider an actively maintained alternative.

Also in this package

Other skills from kevmoo/kevmoo_skills · top by installs.

npx skills add kevmoo/kevmoo_skills

Browse all from kevmoo/kevmoo_skills

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Repository health

Stars 14
Default branch main
Open issues 0
Status Archived

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 5,469 B
  • docs SUMMARY.md 550 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 17 installs

SKILL.md

🧭 Sidequest (/sidequest)

Synthesizes task hierarchies and context drift into a visual session map.

⚠️ Mandatory Execution Contract (5 Core Rules)

  1. Tool-Driven State Updates: Always update state via the CLI tool (sidequest). Never edit sidequest.json manually.
  2. Tool-Driven Map Compilation: Always generate/compile sidequest.md via sidequest. Never format the markdown map by hand.
  3. Strict Internal Privacy: NEVER mention or reference sidequest.json in user-facing conversation. Treat JSON state as a private implementation detail.
  4. Markdown User Interface: Always reference sidequest.md or provide concise inline markdown summaries when communicating progress to the human.
  5. Subagent History Ingestion: Never read transcript.jsonl directly in the main conversation; delegate deep history rebuilds exclusively via /sidequest rebuild.

🏗️ Storage & Architecture

  • Session-Private Artifacts: All state files (sidequest.json, sidequest.md) reside strictly in the session artifact directory (auto-discovered via ANTIGRAVITYCONVERSATIONID, CLAUDEARTIFACTDIR, GEMINIARTIFACTDIR, or --dir). Never write to user repositories or dotfiles.
  • Compaction Resilient: sidequest.json maintains the deterministic state model (quests, completion orders, VCS state, step watermark) across context truncations.

🧭 Hierarchy & Syntax Specification

Level Syntax / Prefix Description Status Indicators
Main Quest Main Quest N: High-level initiatives / chapters ⚔️ [ACTIVE], 🏆 [COMPLETED], ⏸️ [PAUSED]
Sub-Quest Sub-Quest N.M: Planned milestones 🛡️
Blocker Blocker N.M.K: Critical-path unplanned blocker 👾 Active, 💀 Resolved
Step Step N.M.K: Planned action item 👣 Active, 👣 Done
Side Quest [Active] / 🎒 [Parked] Tangents / rabbit holes (G1, S1) 🌿
  • Completion Order ([#N ⭐]): Completed items receive sequential tags ([#1], [#2]). The most recently completed item receives the star ([#N ⭐]).
  • VCS Lifecycle: Track working copy state per quest: 📝 Dirty -> 📦 Local Commit -> 🚀 Uploaded -> 🎉 Merged -> 🧹 Clean.

🚀 Execution Workflow

When /sidequest triggers (via /sidequest, "where are we?", or context drift):

Mode A: In-Session CLI Mutation (Default O(1))

Execute sidequest (or dart run <path-to-skill>/bin/sidequest.dart):

# 1. Inspect Current State (Outputs compact overview to stdout)
sidequest status

# 2. Initialize or Add Quests, Sub-Quests, Steps, Blockers
sidequest init "Title"
sidequest subquest add 1 "UI Implementation"
sidequest step add 1.1 "Draft UI widget"
sidequest blocker add 1.1 "Broken build dependency"
sidequest sidequest add "Tangent item" [--global] [--parked] [--note="..."]

# 3. Batch Operations (Atomic multi-item execution in a single call)
sidequest batch '[{"op":"subquest_add","quest_id":"1","title":"Backend"},{"op":"step_add","subquest_id":"1.2","title":"API client"}]'

# 4. Complete One or Multiple Items (Atomic disk write & star update)
sidequest complete 1.1.1 1.1.2 1.1

# 5. Update VCS Lifecycle
sidequest vcs 1 --stage=dirty|local_commit|uploaded|merged|clean [--branch=B] [--files=F]

# 6. Reopen or Remove
sidequest reopen 1.1
sidequest remove 1.1.2

User Output: Output a brief, punchy chat summary covering active ⚔️ Main Quest, current 🛡️ Sub-Quest, VCS status, and recommended next step. Always place the clickable link to the generated artifact at the very BOTTOM of the chat reply with an emoji anchor so it is easy to find and click:

🗺️ Full Session Map: [sidequest.md](file:///path/to/sidequest.md)

Mode B: Subagent Transcript Rebuild (/sidequest rebuild)

Use only when initializing from long unmapped history or explicitly requested via /sidequest rebuild:

  1. Spawn Auditor Subagent: TypeName: "research", Role: "Sidequest Log Auditor", passing baseline sidequest.json and [auditorprompt.txt](resources/auditorprompt.txt).
  2. Delta Audit: Subagent inspects transcript.jsonl from watermark.stepIndex onwards and returns audited JSON payload in send_message.
  3. Merge & Emit: Parent runs sidequest merge-audit --input=<payload_file> to update JSON and compile sidequest.md.

🤝 Parked Item Escalation

When parking side quests (🎒 [Parked / Tracked for Later]), check available issue trackers and offer:

"Would you like me to file an issue in your project tracker (gh issue create / local tracker) so this parked item survives across sessions?"