メニュー
Load project context efficiently for Conductor workflows. Use when starting work on a track, implementing features, or needing project context without consuming excessive tokens.
デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。
インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
直接コマンドでは確認用 Prompt が省略されます。実行前にソースを確認してください。
npx skills add https://github.com/Ibrahim-3d/orchestrator-supaconductor --skill context-loaderコマンドは1行のまま表示されます。コピー前に横へスクロールして全体を確認してください。
ローカルで確認しますか？SkillsMP が現在取得できるファイルをダウンロードできます。
Master coordinator for the Evaluate-Loop workflow v3. Supports GOAL-DRIVEN entry, PARALLEL execution via worker agents, BOARD OF DIRECTORS deliberation, and message bus coordination. Dispatches specialized workers dynamically, monitors via message bus, aggregates results. Uses metadata.json v3 for parallel state tracking. Use when: '/go <goal>', '/conductor implement', 'start track', 'run the loop', 'orchestrate', 'automate track'.
File-based message queue for inter-agent coordination. Used by workers AND board directors to communicate. Provides: progress updates, task completion signals, file locking, board deliberation. Core infrastructure for parallel execution.
Loads relevant patterns and known errors before track planning. Searches conductor/knowledge/ for solutions we've used before and errors we've encountered. Injects findings into the planner prompt to prevent reinventing solutions and repeating mistakes. Triggered automatically by orchestrator before PLAN step.
SOC 職業分類に基づく
SKILL.md を表示中
| name | context-loader |
| description | Load project context efficiently for Conductor workflows. Use when starting work on a track, implementing features, or needing project context without consuming excessive tokens. |
Efficiently load and manage project context for Conductor's context-driven development workflow.
Use this skill when:
/orchestrator-supaconductor:implement workflowBefore scanning files, check for:
.claudeignore - Claude-specific ignores.gitignore - Standard git ignores# Check for ignore files
ls -la .claudeignore .gitignore 2>/dev/null
Use git for tracked files:
git ls-files --exclude-standard -co | head -100
For directory structure:
git ls-files --exclude-standard -co | xargs -n 1 dirname | sort -u
| Priority | File Type | Examples |
|---|---|---|
| 1 | Manifests | package.json, Cargo.toml, pyproject.toml |
| 2 | Conductor | conductor/product.md, conductor/tech-stack.md |
| 3 | Track | conductor/tracks/<id>/spec.md, plan.md |
| 4 | Config | tsconfig.json, .env.example |
For files over 1MB:
ls -la before reading
500KB: read_file first 20 + last 20 lines only
1MB: Skip entirely, log as "skipped: too large"
1. Load CLAUDE.md (if exists)
2. Load conductor/product.md (project vision)
3. Load conductor/tech-stack.md (technical context)
4. Load conductor/tracks.md (completed work — prevents duplicate effort)
5. Load current track spec.md (requirements)
6. Load current track plan.md (tasks — check [x] vs [ ] status)
This skill is used by multiple loop agents:
loop-planner — loads context before creating a planloop-executor — loads context before implementing tasksconductor-orchestrator — loads context to determine current loop stepCritical: Always load tracks.md and check plan.md task markers to prevent duplicate work across sessions.
After loading context, summarize:
## Project Context Loaded
**Product**: [one-line summary]
**Tech Stack**: [key technologies]
**Current Track**: [track name/id]
**Active Phase**: [current phase]
**Pending Tasks**: [count of [ ] tasks]
**Completed Tasks**: [count of [x] tasks]
**Loop Step**: [current Evaluate-Loop step — Plan/Execute/Evaluate/Fix]