agentleFS
AbdulSamad94AbdulSamad94/Personal-AI-Employee/.agents/skills/process_tasks/SKILL.md

process-tasks

Skill8 starsChanged 4 months agoMITView on GitHub

Process all pending task files in vault/Needs_Action/. Use this skill whenever new .md files appear in Needs_Action/, when the user says "process tasks", "check inbox", "what needs attention", or "handle pending items". Also auto-triggers every 30 minutes during scheduled runs.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

24 rules. About 0 are common, 19 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈19

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Never skip a file. Never act without reading Company_Handbook.md first.

    line 14
  2. Distinctive:

    Before doing ANYTHING else, read these two files:

    line 18
  3. Distinctive:

    Read the frontmatter (type, from, subject, status)

    line 28
  4. Distinctive:

    Read the full content

    line 29
  5. Distinctive:

    If Needs_Action/ is empty, update Dashboard.md with: "Last checked: [timestamp] — No pending tasks" Then stop.

    line 32
  6. Very specific:

    Create approval file in vault/Pending_Approval/

    line 103
  7. Very specific:

    DO NOT send anything until approval file moves to vault/Approved/

    line 104
  8. Distinctive:

    NEVER process payments automatically

    line 116
  9. Distinctive:

    ALWAYS create approval file regardless of amount

    line 117
  10. Distinctive:

    Wait for explicit human approval

    line 119
  11. Distinctive:

    Create approval file with expiry of 2 hours instead of 24

    line 124
  12. Distinctive:

    Add to Dashboard.md under "🚨 Urgent Items"

    line 125
  13. Very specific:

    After processing ALL tasks, update vault/Dashboard.md:

    line 129
  14. Distinctive:

    Add each action to "Recent Activity" with timestamp

    line 134
  15. Distinctive:

    Update "Last Updated" timestamp

    line 135
  16. Distinctive:

    When a task is fully handled:

    line 154
  17. Distinctive:

    Do NOT delete any files — always move to Done/

    line 158
  18. Very specific:

    NEVER send an email without an approval file in vault/Approved/

    line 162
  19. Very specific:

    NEVER post to LinkedIn without an approval file in vault/Approved/

    line 163
  20. Distinctive:

    NEVER process any payment automatically

    line 164
  21. Distinctive:

    NEVER delete files — always move to Done/

    line 165
  22. Distinctive:

    ALWAYS follow Company_Handbook.md tone for any draft content

    line 166
  23. Distinctive:

    ALWAYS log every action taken

    line 167
  24. Distinctive:

    ALWAYS update Dashboard.md after every session

    line 168

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Skill: Process Tasks

Purpose

You are a Personal AI Employee. Your job is to process every pending task file in vault/Needs_Action/ and take the appropriate action for each one. Never skip a file. Never act without reading Company_Handbook.md first.

Step 1 — Read Your Rules First

Before doing ANYTHING else, read these two files:

  • vault/Company_Handbook.md — your rules of engagement
  • vault/Business_Goals.md — your business context

Step 2 — Scan Needs_Action

List all .md files in vault/Needs_Action/. For each file:

  • Read the frontmatter (type, from, subject, status)
  • Read the full content
  • Determine the task type: email / whatsapp / file_drop / payment / linkedin

If Needs_Action/ is empty, update Dashboard.md with: "Last checked: [timestamp] — No pending tasks" Then stop.

Step 3 — Classify Each Task

For every task file, classify it as one of:

ClassificationCriteria
SAFEInternal only, no external action needed
NEEDS_REPLYRequires sending an email or WhatsApp response
NEEDS_POSTRequires posting to LinkedIn
NEEDS_PAYMENTInvolves any money or invoice
URGENTContains words: urgent, asap, emergency, critical

Step 4 — Create Plan.md

For EVERY task, create a plan file at: vault/Plans/PLAN_[task_name]_[YYYYMMDD_HHMMSS].md

Use this exact format:

created: [ISO timestamp] task_source: [original filename] task_type: [email/whatsapp/file_drop/payment] classification: [SAFE/NEEDS_REPLY/NEEDS_POST/NEEDS_PAYMENT/URGENT] status: in_progress requires_approval: [yes/no]


Objective

[One sentence: what needs to happen]

Context

[Who sent it, what they want, any relevant background]

Steps

  • Read task file
  • Classified as: [type]
  • [Next specific action]
  • [Action after that]
  • Request approval (if needed)
  • Execute or wait for approval
  • Update Dashboard.md
  • Move task to Done/
  • Log action

Decision Reasoning

[Why you chose this plan. What rules from Company_Handbook.md apply.]

Risk Assessment

[Low / Medium / High — and why]

Step 5 — Act Based on Classification

For SAFE tasks:

  • Handle directly (rename, summarize, organize)
  • No approval needed
  • Move task file to vault/Done/ when complete

For NEEDS_REPLY tasks:

  • Draft the full reply text
  • Create approval file in vault/Pending_Approval/
  • DO NOT send anything until approval file moves to vault/Approved/
  • See hitl-workflow skill for exact approval file format

For NEEDS_POST tasks:

  • Draft the LinkedIn post (max 1300 characters)
  • Make it professional and value-driven
  • Create approval file in vault/Pending_Approval/
  • See linkedin-skill for post format

For NEEDS_PAYMENT tasks:

  • NEVER process payments automatically
  • ALWAYS create approval file regardless of amount
  • Flag in Dashboard.md under "⚠️ Payment Pending"
  • Wait for explicit human approval

For URGENT tasks:

  • Process immediately before all other tasks
  • Create approval file with expiry of 2 hours instead of 24
  • Add to Dashboard.md under "🚨 Urgent Items"

Step 6 — Update Dashboard.md

After processing ALL tasks, update vault/Dashboard.md:

  • Increment "Active Tasks" count for items in Needs_Action/
  • Increment "Pending Approval" count for items in Pending_Approval/
  • Increment "Completed Today" count for items moved to Done/
  • Add each action to "Recent Activity" with timestamp
  • Update "Last Updated" timestamp

Step 7 — Log Everything

Append to vault/Logs/[YYYY-MM-DD].md:

[ISO timestamp] | [action_type] | [task_file] | [result] | [approval_required]

Example:

2026-01-07T14:30:00Z | email_draft | EMAIL_abc123.md | approval_requested | yes
2026-01-07T14:31:00Z | file_processed | FILE_report.md | moved_to_done | no

Step 8 — Move Completed Tasks

When a task is fully handled:

  • Move task .md file from vault/Needs_Action/ to vault/Done/
  • Mark Plan.md status as: complete
  • Do NOT delete any files — always move to Done/

Rules (NEVER Break These)

  • NEVER send an email without an approval file in vault/Approved/
  • NEVER post to LinkedIn without an approval file in vault/Approved/
  • NEVER process any payment automatically
  • NEVER delete files — always move to Done/
  • ALWAYS follow Company_Handbook.md tone for any draft content
  • ALWAYS log every action taken
  • ALWAYS update Dashboard.md after every session

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.