Rechercher

PythonReactSEOPDF
OpenClaw

focus-tracker

Maintain persistent focus on active work across compactions and session restarts. Use when starting a multi-step project, when resuming after compaction, when the user says "we're working on X", or when you notice you've lost track of what was being done. Also use proactively at session start to check if there's unfinished work. Prevents the agent from forgetting mid-project context, drifting between topics, or asking "what were we doing?" after a compaction.

crewhaus0 installations selon la source☆ 0 Étoiles du dépôtv1.0.0

Installation

Suivez les instructions du dépôt pour votre agent. Consultez les fichiers et dépendances avant l’installation.

SkillHub CLI

skillhub install crewhaus/focus-tracker

Examinez la source et la compatibilité, puis suivez les étapes d’installation.

Donnez le lien à votre agent et demandez-lui d’examiner instructions et dépendances avant d’expliquer les étapes.

Voir la source ↗Guides

Instructions

Focus Tracker

Maintain a FOCUS.md file in the workspace root as persistent working state.

When to Read FOCUS.md

  • Every session start (before responding to first message)
  • After every compaction (if context feels thin, read it)
  • When the user asks about current work ("what are we doing?", "where were we?")

When to Write/Update FOCUS.md

  • When starting a new focused project — create or replace the active focus
  • When a sub-task completes — check off items, update status
  • When priorities shift — user redirects, update immediately
  • When work is DONE — archive to FOCUS-LOG.md and clear FOCUS.md
  • Before compaction memory flushes — ensure FOCUS.md reflects current state

FOCUS.md Format

# FOCUS — [Project Name]
**Started:** YYYY-MM-DD HH:MM UTC
**Status:** active | paused | blocked
**Context:** One-line summary of WHY we're doing this

## Objective
What we're trying to accomplish (1-3 sentences max)

## Plan
Numbered steps. Check off as completed.
- [x] Step 1 — done
- [ ] Step 2 — in progress (note current state)
- [ ] Step 3 — next

## Active State
What's happening RIGHT NOW. What was the last thing done, what's the next action.
This is the most important section — it's the "resume point" after compaction.

## Blockers
Anything waiting on the user, external systems, or decisions.

## Sub-Agents
Any spawned sub-agents and their status.
- `label` — task description — status (running/done/failed)

Rules

  1. One focus at a time. If the user pivots, archive the old focus first.
  2. Keep it short. FOCUS.md should be <50 lines. It's a resume point, not a journal.
  3. Active State is sacred. Always update it before stopping work. Future-you reads this first.
  4. Don't duplicate memory files. FOCUS.md tracks what we're doing now. Daily memory files track what happened. Different jobs.
  5. Archive when done. Append completed focuses to FOCUS-LOG.md with completion date and outcome,...

Informations sur la source

Source
github
Repository
openclaw/skills
Path
skills/crewhaus/focus-tracker
Version
1.0.0
Commit
8c199af950fa0bd6ebca38b05dc7d01d7ffe06cc
Source de classification
rules
Statut
active
Visibilité
public

Ces informations proviennent de l’analyse indexée et peuvent être anciennes. Vérifiez les fichiers actuels.

Contenu extrait

SKILL.md
# Focus Tracker

Maintain a `FOCUS.md` file in the workspace root as persistent working state.

## When to Read FOCUS.md

- **Every session start** (before responding to first message)
- **After every compaction** (if context feels thin, read it)
- **When the user asks about current work** ("what are we doing?", "where were we?")

## When to Write/Update FOCUS.md

- **When starting a new focused project** — create or replace the active focus
- **When a sub-task completes** — check off items, update status
- **When priorities shift** — user redirects, update immediately
- **When work is DONE** — archive to `FOCUS-LOG.md` and clear FOCUS.md
- **Before compaction memory flushes** — ensure FOCUS.md reflects current state

## FOCUS.md Format

```markdown
# FOCUS — [Project Name]
**Started:** YYYY-MM-DD HH:MM UTC
**Status:** active | paused | blocked
**Context:** One-line summary of WHY we're doing this

## Objective
What we're trying to accomplish (1-3 sentences max)

## Plan
Numbered steps. Check off as completed.
- [x] Step 1 — done
- [ ] Step 2 — in progress (note current state)
- [ ] Step 3 — next

## Active State
What's happening RIGHT NOW. What was the last thing done, what's the next action.
This is the most important section — it's the "resume point" after compaction.

## Blockers
Anything waiting on the user, external systems, or decisions.

## Sub-Agents
Any spawned sub-agents and their status.
- `label` — task description — status (running/done/failed)
```

## Rules

1. **One focus at a time.** If the user pivots, archive the old focus first.
2. **Keep it short.** FOCUS.md should be <50 lines. It's a resume point, not a journal.
3. **Active State is sacred.** Always update it before stopping work. Future-you reads this first.
4. **Don't duplicate memory files.** FOCUS.md tracks *what we're doing now*. Daily memory files track *what happened*. Different jobs.
5. **Archive when done.** Append completed focuses to `FOCUS-LOG.md` with completion date and outcome,...

Skills associés

OpenClaw

x402hub

capncoconut

Register, communicate, and earn on the x402hub AI agent marketplace. Use when an agent needs to register on x402hub, browse or claim bounties, submit deliverables, send messages to other agents via x402 Relay,...

OpenClaw

claw

capevace

Real-time event bus for AI agents. Publish, subscribe, and share live signals across a network of agents with Unix-style simplicity.

OpenClaw

captchas-openclaw

captchasco

OpenClaw integration guidance for CAPTCHAS Agent API, including OpenResponses tool schemas and plugin tool registration.