fur-status

star 1

Show a concise Fur project status using task counts, latest progress snapshot, ready tasks, tracker sync state, and the next recommended action.

0furkancolak By 0furkancolak schedule Updated 6/10/2026

name: fur-status skill_class: orchestrator skill_version: 2 default_response_depth: concise description: >- Show a concise Fur project status using task counts, latest progress snapshot, ready tasks, tracker sync state, and the next recommended action. disable-model-invocation: true requires: - project_root - fur_planning_dir optional: - state_json - progress_latest - fur_workspace_config quality_contract: must_map_every_ac: false must_report_assumptions: false must_report_verification_truthfully: true must_call_out_risks: false must_include_user_facing_explanation: true self_check_required: true

fur-status

Orients humans and agents in <30 seconds without rereading old task bodies.

Identity

You are a project coordinator. Your job is to give a fast, accurate orientation and one concrete next action.

Goal

Summarize queue depth, latest fur refresh snapshot, workspace registration, and one concrete next action.

When to Use

  • "Where are we?", "what's next?", "what's in ready?"
  • Right after closing a task to pick the next unit of work.
  • Long session: quick re-ground before more edits.

When NOT to Use

  • Need to create or reshape tasks from raw notes — this skill is read-only.
  • Need implementation — status does not write code.
  • Need verification or task closure — status does not move tasks or run gates.

Context Loading Contract

Load in this order:

  1. docs/ai/config.json.
  2. docs/ai/state.json when present.
  3. tasks/{backlog,ready,done} counts.
  4. progress/latest.md head (~80-120 lines).
  5. .fur.workspace/config.json if present.

Do not load full task bodies or archived snapshots.

Workflow

Phase 1: Read planning state (chat-first)

  1. Count *.md in tasks/{backlog,ready,done} and list plans/*.md.
  2. For each plan file: read manifest + reconcile task folder statuses; render one compact ## Plan summary line per plan.
  3. Do not stop at "run fur plan status" — the user interacts via AI; the dashboard must appear in this message.
  4. Optional shell: fur progress / fur plan status for cross-check only.
  5. If docs/ai is missing, report "run fur init" and stop.

Phase 2: Read latest markdown snapshot

  1. If progress/latest.md exists, read only the head (~80–120 lines): git snippet, counts, tracker hints, newest ready filenames.
  2. Do not load archived snapshots or full done/ histories unless the user asked for archaeology.

Phase 3: Workspace awareness

  1. If parent .fur.workspace/config.json exists, note defaultTaskSource, write policy, and whether this repo path appears under repositories[].
  2. If not registered, the next external import/write may be blocked — say so plainly.

Phase 4: Recommend a single next step

Pick exactly one concrete next action: name a ready task file, run fur refresh / fur compact / fur workspace doctor, run fur init, or state what is blocked.

Phase 5: Self-review

Before finalizing, verify:

  • Did I provide counts for backlog, ready, and done?
  • Did I read only the head of progress/latest.md?
  • Did I recommend exactly one next action?
  • Did I match the output contract for this skill class?

If any answer is no, continue working before responding.

Rules

  • Read-only: no file mutations in this skill.
  • Never claim work is complete without matching task location + verification trail.
  • Prefer paths and counts over pasting markdown.
  • Next Action = one short plain-text sentence (or caveman fragment if responseDepth: concise or user invoked caveman). Never append YAML (status, next_skill).
  • If user attached superpowers or caveman skills, follow those for tone/process; this skill stays read-only.
  • If progress/ is huge, suggest fur compact and mention FUR_PROGRESS_KEEP override (default keep is 8).
  • Do not trigger external methodology from this skill.

Output

Plain markdown only. End with Next Action as prose — no YAML block.

## Summary

- Backlog: [n]
- Ready: [n]
- Done: [n]
- Plans: [n]
- Workspace: found | missing
- Repo registered: yes [id] | no

## Plan summary

[One compact line per active plan: `slug`: percent (done/total) · next: task or complete.]

## Latest Snapshot

[2–5 bullets distilled from progress/latest.md]

## Next Action

[One short sentence. Example: CI yeşil olunca merge; sonra Epic 7 audit.]

Anti-patterns

  • Do not dump every task ever created.
  • Do not recommend more than one next action.
  • Do not mutate files in this skill.
  • Do not load full archived snapshots.

Examples

Reference examples:

  • ../_shared/examples/orchestrator-example.md
  • ../_shared/anti-patterns/global.md
  • ../_shared/anti-patterns/orchestrator.md

Use these examples to calibrate response depth, evidence quality, output structure, and self-check behavior.

Install via CLI
npx skills add https://github.com/0furkancolak/fur-skills --skill fur-status
Repository Details
star Stars 1
call_split Forks 0
navigation Branch main
article Path SKILL.md
More from Creator
0furkancolak
0furkancolak Explore all skills →