agentleFS
agenticlukeagenticluke/live-message-ops-plus/skill/SKILL.md

messages-ops

Skill0 starsChanged 48 days agoMITView on GitHub

Evidence-first steps for ECC live messages. Use when the user wants to read texts or direct messages, find a recent one-time code, check a thread before a reply, or show which message source was checked.

6 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions6 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 6

    1. Fmarzochi/EGC★ 61same name
    2. gongyijie85/dsh-ecc★ 8same name
    3. Rojios/everything-claude-code★ 1same name
    4. SamuelNDCE/claude-super-skill-library★ 1same name
    5. davidanthonyn/claude-skills★ 0same name
    6. renepf/claude-skills★ 0same 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.

The file

Show its rules and text

We split this skill into the rules it gives an agent

40 rules. 0 are common, 39 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 39

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill for live message tasks.

    line 11
  2. Distinctive:

    Use these ECC skills when they fit:

    line 17
  3. Distinctive:

    Do not switch skills only because a message names an email address.

    line 24
  4. Distinctive:

    Use it when the user asks to:

    line 28
  5. Distinctive:

    Read messages, texts, or direct messages

    line 30
  6. Distinctive:

    Check a thread before writing a reply

    line 32
  7. Distinctive:

    Use only the source the user asked for.

    line 38
  8. Distinctive:

    If the source is not clear, ask which app or service to check.

    line 39
  9. Distinctive:

    Use an approved helper or normal access path when one exists.

    line 40
  10. Distinctive:

    Do not open raw message files or message databases unless the user clearly asks and no safe helper exists.

    line 41
  11. Distinctive:

    Read only the messages needed for the task.

    line 42
  12. Distinctive:

    Treat one-time codes as private.

    line 43
  13. Distinctive:

    Do not send a reply unless the user clearly asks you to send it.

    line 44
  14. Distinctive:

    If the user asks for a draft, draft it but do not send it.

    line 45
  15. Distinctive:

    Do not mark a message as read, delete it, archive it, or change the thread unless asked.

    line 46
  16. Distinctive:

    If sign-in or two-step checks block access, stop and state the exact block.

    line 47
  17. Distinctive:

    Do not keep trying the same blocked sign-in path.

    line 48
  18. Distinctive:

    Before you act, find:

    line 54
  19. Distinctive:

    If one detail is missing but the task is still clear, use the smallest safe scope.

    line 61
  20. Distinctive:

    If the task may lead to a reply:

    line 65
  21. Distinctive:

    Read the newest incoming message.

    line 67
  22. Distinctive:

    Read enough earlier messages to understand it.

    line 68
  23. Distinctive:

    Do not guess what the thread says.

    line 73
  24. Distinctive:

    Search recent local messages first.

    line 79
  25. Distinctive:

    Check the newest matching message first.

    line 81
  26. Distinctive:

    Make sure the code is recent and matches the named service.

    line 82
  27. Distinctive:

    Stop when the code is found or the narrow search is done.

    line 83
  28. Distinctive:

    If two codes match, return the newest one and note that an older code also exists.

    line 85
  29. Very specific:

    Never use Sent unless the send action was confirmed.

    line 103
  30. Distinctive:

    If access is blocked:

    line 152
  31. Distinctive:

    Do not treat email as a text or direct message.

    line 173
  32. Distinctive:

    Do not claim a search happened when the source was not known.

    line 174
  33. Distinctive:

    Do not run a wide search for a recent code.

    line 175
  34. Distinctive:

    Do not show extra private messages when one result is enough.

    line 176
  35. Distinctive:

    Do not use an old code when a newer code exists.

    line 177
  36. Distinctive:

    Do not draft or send a reply before reading the latest thread.

    line 178
  37. Distinctive:

    Do not hide a sign-in or access block.

    line 179
  38. Distinctive:

    Do not say a reply was sent without proof.

    line 180
  39. Distinctive:

    Before you answer, make sure:

    line 184
  40. Distinctive:

    No message was sent or changed without clear user approval.

    line 191

“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “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

Message Operations

Created by ECC. Credit to the original author and project.

Use this skill for live message tasks. This includes texts, direct messages, recent one-time codes, and thread checks before the next step.

This skill is not for email. Use email-ops when the main source is an inbox.

Related Skills

Use these ECC skills when they fit:

  • email-ops when the source is email
  • connections-optimizer when a direct message is part of outreach
  • lead-intelligence when a live thread should guide who to contact or how to start
  • knowledge-ops when the user asks to save useful thread details for later

Do not switch skills only because a message names an email address. Check where the message was received.

When to Use This Skill

Use it when the user asks to:

  • Read messages, texts, or direct messages
  • Find a recent sign-in or one-time code
  • Check a thread before writing a reply
  • Name the source, sender, thread, or time checked

Safety Rules

  • Name the source before you claim that you checked it.
  • Use only the source the user asked for.
  • If the source is not clear, ask which app or service to check.
  • Use an approved helper or normal access path when one exists.
  • Do not open raw message files or message databases unless the user clearly asks and no safe helper exists.
  • Read only the messages needed for the task.
  • Treat one-time codes as private. Return only the needed code and basic source details.
  • Do not send a reply unless the user clearly asks you to send it.
  • If the user asks for a draft, draft it but do not send it.
  • Do not mark a message as read, delete it, archive it, or change the thread unless asked.
  • If sign-in or two-step checks block access, stop and state the exact block.
  • Do not keep trying the same blocked sign-in path.

Steps

1. Set the Scope

Before you act, find:

  • The message app or service
  • The sender, receiver, thread, or service name
  • The time range
  • The task type: find, read, check, draft, or send

If one detail is missing but the task is still clear, use the smallest safe scope. For example, search recent texts from the named bank.

2. Read Before You Draft

If the task may lead to a reply:

  1. Read the newest incoming message.
  2. Read enough earlier messages to understand it.
  3. Note any open question, promise, date, or next step.
  4. Draft a reply only if asked.
  5. Hand off to a related skill only when it fits the task.

Do not guess what the thread says.

3. Find One-Time Codes Fast

For a one-time code:

  1. Search recent local messages first.
  2. Limit the search by service or sender when known.
  3. Check the newest matching message first.
  4. Make sure the code is recent and matches the named service.
  5. Stop when the code is found or the narrow search is done.

If two codes match, return the newest one and note that an older code also exists. Do not assume an old code still works.

4. Report the Evidence

State:

  • The source you checked
  • The sender, thread, or service, when known
  • The time range you checked
  • A short result
  • One exact status:
    • Read
    • Code found
    • Not found
    • Blocked
    • Draft ready
    • Sent

Never use Sent unless the send action was confirmed.

Output Format

Source
- App or service:
- Sender, thread, or service:
- Time range:

Result
- Short message summary or one-time code:

Status
- Read / Code found / Not found / Blocked / Draft ready / Sent

For a blocked task, add:

Block
- Exact reason:
- What the user must do next:

Example

User request:

Find the latest login code from Acme in my texts.

Good result:

Source
- App or service: Local text messages
- Sender, thread, or service: Acme
- Time range: Last 15 minutes

Result
- One-time code: 482193
- Received: 2 minutes ago

Status
- Code found

If access is blocked:

Source
- App or service: Local text messages
- Sender, thread, or service: Acme
- Time range: Last 15 minutes

Result
- I could not read the thread.

Status
- Blocked

Block
- Exact reason: The message app needs sign-in approval.
- What the user must do next: Approve the sign-in, then ask me to try again.

Common Mistakes

  • Do not treat email as a text or direct message.
  • Do not claim a search happened when the source was not known.
  • Do not run a wide search for a recent code.
  • Do not show extra private messages when one result is enough.
  • Do not use an old code when a newer code exists.
  • Do not draft or send a reply before reading the latest thread.
  • Do not hide a sign-in or access block.
  • Do not say a reply was sent without proof.

Final Check

Before you answer, make sure:

  • The message source is named.
  • The sender, service, thread, or exact block is named.
  • The time range is clear.
  • The result does not expose extra private data.
  • The status is exact.
  • No message was sent or changed without clear user approval.

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.