Install
$ agentstack add skill-fractalmind-ai-okr-manager-skill-okr-manager-skill ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
OKR Manager
A skill for AI agents to manage OKR (Objectives and Key Results) lifecycle — from creation to completion.
When to Use
- Your human assigns a new objective or goal
- Creating or restructuring an OKR in your OKR file
- Reviewing existing OKRs for completeness
- Heartbeat/periodic audit: checking if any OKR lacks completion standards
- Reporting OKR status to your human or team
Configuration
This skill works with any markdown-based OKR file. Set these in your workspace:
| Config | Default | Description | |--------|---------|-------------| | OKR file path | OKR.md | Where OKRs are stored | | Language | en | Primary language for OKR content (en, zh, etc.) | | Status markers | PENDING / IN PROGRESS / COMPLETE | KR status labels | | Priority markers | P0 / P1 / P2 | Priority levels |
OKR Creation Checklist
Every OKR MUST have ALL of these. Reject or flag any OKR missing items.
1. Objective (1 sentence)
- Starts with a verb (Build / Implement / Complete / Deliver)
- Describes the outcome, not the activity
- Bad: "Do orderbook development" / Good: "Complete orderbook development, testing, and launch"
2. Success Criteria (1-3 bullets, quantifiable)
- Must be binary verifiable: done or not done
- Must include at least ONE measurable metric
- Template: "Complete {N} {specific actions} in {environment}, verify {specific metric}"
- Examples:
- "Complete 1 gasless transaction on TestNet"
- "Third-party PR 24h first-response rate >= 95%"
- "E2E test pass rate 100%, matching latency Dependency order: KR1 → KR2 → ... → KRn
KR1: {title} — {status}
- Deliverable: {what is produced}
- Outcome: {observable change when KR is complete}
- Verification: {how to prove done}
KR2: {title} — {status}
- Deliverable: {what is produced}
- Depends on: KR1
- Outcome: {observable change when KR is complete}
- Verification: {how to prove done}
Tasks / Milestones (recommended)
- [ ] {execution step tied to KR1 / owner / dependency}
- [ ] {execution step tied to KR2 / owner / dependency}
## OKR Lifecycle
Human assigns objective ↓ [Create OKR] → Use this skill to ensure checklist passes ↓ [Write to OKR file] → Confirm Success Criteria + KRs are complete ↓ [Human confirms] → If issues, iterate and fix ↓ [Execute & Track] → Each heartbeat/check-in: review progress, push forward ↓ [KR Complete] → Update status, notify human ↓ [Success Criteria met] → Mark OKR as ACHIEVED
## Operations
### Creating an OKR
1. Parse the human's objective into a clear, verb-first Objective statement
2. Draft measurable Success Criteria (what does "done" look like?)
3. Draft sequential, result-oriented KRs with dependencies
4. Add Tasks / Milestones if execution needs staging, owner mapping, or unblock steps
5. Validate against the checklist above
6. Write to OKR file
7. Confirm with human
### Updating KR Status
When a KR completes:
1. Update status: `PENDING` → `IN PROGRESS` → `✅ COMPLETE`
2. Add completion evidence (PR number, CI link, test results)
3. Check if downstream KRs are now unblocked
4. If all KRs complete → check Success Criteria → mark OKR ACHIEVED
### Heartbeat OKR Audit
During periodic check-ins:
1. Read the OKR file
2. For each ACTIVE OKR:
- What KR is currently in progress?
- Is anything blocked? For how long?
- What's the next action?
3. Report summary to human (only if there are changes or blockers)
4. Push forward: assign tasks, trigger CI, follow up on blockers
### Reporting Status
Format for status reports:
[OKR Status] • {OKR Name}: KR{n}/{total} — {current KR title} ({status}) • {OKR Name}: KR{n}/{total} — {current KR title} ({status}) Blockers: {list or "none"}
## Priority Emojis
| Priority | Emoji | Meaning |
|----------|-------|---------|
| P0 | 🔴 | Must do — blocks everything else |
| P1 | 🟡 | Important — do this week |
| P2 | 🟢 | Nice to have — do when capacity allows |
## Common Anti-patterns (Avoid)
1. **Vague completion standards**: "launch" "improve" "optimize" — must quantify
2. **Missing verification**: KR says what to do but not how to prove it's done
3. **Task masquerading as KR**: "build agent manager" / "integrate Slack" — move this into Tasks / Milestones
4. **KR too large**: Single KR contains multiple independent pieces of work → split it
5. **No priority**: All OKRs appear equally important → must rank
6. **Passive voice**: "wait for deploy" "wait for confirmation" → change to active: "trigger deploy" "verify completion"
7. **No owner**: Nobody is accountable → every OKR needs an owner
8. **Stale OKRs**: OKR status not updated for >1 week → audit and update or archive
## Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- **Author:** [fractalmind-ai](https://github.com/fractalmind-ai)
- **Source:** [fractalmind-ai/okr-manager-skill](https://github.com/fractalmind-ai/okr-manager-skill)
- **License:** MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.