reasoning-loop
Skill8 starsChanged 4 months agoMITView on GitHub
Deep reasoning loop for complex multi-step tasks. Use when a task requires planning across multiple files, cross-referencing business goals, making decisions with multiple options, or when the user says "think through this", "make a plan", "figure out what to do", or "analyze my tasks".
1 other project uses a version of this.
Adopted by
Most recently taken up by
- omerfarooq223/personal-ai-employeetheir own version, first seen Mar 19, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- omerfarooq223/personal-ai-employee★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository27 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- ceo-briefingSkill
- dashboard-updateSkill
- email-skillSkill
- fetch-library-docsSkill
- hitl-workflowSkill
- linkedin-skillSkill
- mcp-builderSkill
- odoo-accountingSkill
- process-tasksSkill
- reasoning-loopSkill
- skill-creator-proSkill
- skill-validatorSkill
- social-managerSkill
- ceo-briefingSkill
- dashboard-updateSkill
- email-skillSkill
- fetch-library-docsSkill
- hitl-workflowSkill
- linkedin-skillSkill
- mcp-builderSkill
- odoo-accountingSkill
- process-tasksSkill
- skill-creator-proSkill
- skill-validatorSkill
- social-managerSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 22
Read in this exact order:
- Very specific: line 27
All files in vault/Needs_Action/
- Distinctive: line 31
Write a one-paragraph summary of what you observed.
- Distinctive: line 32
Do not form conclusions yet.
- Distinctive: line 38
Ask yourself these questions and write answers:
- Distinctive: line 58
Document your decision and reasoning for each task.
- Very specific: line 137
Create an ESCALATE file in vault/Pending_Approval/ immediately if:
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Skill: Reasoning Loop
Purpose
This skill gives you a structured thinking process for complex tasks that require more than simple file processing. Use this when you need to reason carefully before acting.
The OODA Loop (Observe → Orient → Decide → Act)
Phase 1 — OBSERVE
Gather all relevant information before forming any opinion:
Read in this exact order:
- vault/Company_Handbook.md
- vault/Business_Goals.md
- vault/Dashboard.md (current state)
- All files in vault/Needs_Action/
- Most recent 3 files in vault/Done/ (for context on recent work)
- Most recent briefing in vault/Briefings/ (if exists)
Write a one-paragraph summary of what you observed. Do not form conclusions yet.
Phase 2 — ORIENT
Analyze what you observed against your goals:
Ask yourself these questions and write answers:
- What is the most urgent thing that needs attention right now?
- What aligns with the goals in Business_Goals.md?
- What would be risky to do without human approval?
- What can I handle completely on my own?
- Are there any patterns? (e.g., same client contacting multiple times)
- Are there any deadlines I should know about?
Phase 3 — DECIDE
For each task in Needs_Action/, decide ONE of:
| Decision | When to Use |
| ACT_NOW | Safe, internal, no external action |
| DRAFT_AND_APPROVE | Needs external action (email/post/payment) |
| ESCALATE | Too complex, risky, or ambiguous — flag for human |
| DEFER | Low priority, can wait for next cycle |
Document your decision and reasoning for each task.
Phase 4 — ACT
Execute decisions in this priority order:
- URGENT items first
- ESCALATE items (create flags immediately)
- ACT_NOW items (handle directly)
- DRAFT_AND_APPROVE items (create drafts + approval files)
- DEFER items (add note to Dashboard.md)
Creating Thorough Plan.md Files
For complex tasks, your Plan.md should include:
created: [timestamp] task_source: [filename] reasoning_depth: deep status: in_progress
Situation Analysis
[2-3 sentences describing the full situation]
Stakeholders
- Who sent this?
- Who else is affected?
- Who needs to approve?
Options Considered
Option A: [name]
- Pros: ...
- Cons: ...
- Risk: Low/Medium/High
Option B: [name]
- Pros: ...
- Cons: ...
- Risk: Low/Medium/High
Chosen Approach
[Which option and why]
Execution Steps
- Step 1
- Step 2
- Step 3
Success Criteria
[How will you know this task is complete?]
Fallback Plan
[What to do if the primary approach fails]
Cross-References
- Related tasks: [list any related files in Needs_Action/ or Done/]
- Relevant handbook rules: [which rules apply]
- Business goal alignment: [which goal this serves]
Completion Summary
[Fill this in when done: what actually happened]
When to Stop and Escalate
Create an ESCALATE file in vault/Pending_Approval/ immediately if:
- You are uncertain about the right action
- The task involves legal, medical, or financial edge cases
- The task involves a new client you have no history with
- The task involves credentials or sensitive personal data
- Conflicting instructions exist between handbook and the request
- Any irreversible action with significant consequences
Escalation file format:
type: escalation reason: [why you are escalating] task_source: [original file] your_analysis: [what you figured out so far] question_for_human: [the specific question you need answered]
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

