agentleFS
moltbotmoltbot/molthub/.agents/skills/clawhub-ban-appeals/SKILL.md

clawhub-ban-appeals

Skill9.6k starsChanged 14 days agoMITView on GitHub

Review ClawHub account-ban appeals, distinguish bans from moderation holds or blocked releases, and prepare an evidence-backed moderation decision. Route validated scanner inaccuracies to clawhub-clawscan.

2 other projects use a version of this; 2 new projects in the last 4 weeks, against 0 the 4 before.

Adopted by

Surging: 2 new projects took it in the last 4 weeks, none the 4 weeks before.thin data

Week of Sep 20: 0 new projectsWeek of Sep 27: 0 new projectsWeek of Oct 4: 0 new projectsWeek of Oct 11: 0 new projectsWeek of Oct 18: 0 new projectsWeek of Oct 25: 0 new projectsWeek of Nov 1: 2 new projectsWeek of Nov 8: 0 new projects
Sep 2026Oct 2026Nov 2026
New projects taking a version of this file each week, over the 8 weeks our latest run read: only 2 in all, so this is a hint, not a trend.

Most recently taken up by

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 6 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈6

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Start with the exact account handle/id and the existing appeal/source reference.

    line 16
  2. Distinctive:

    Read the complete appeal thread and prior decisions.

    line 18
  3. Very specific:

    Use available authenticated staff reads to establish the account's current ban or hold, recorded reason/time/provenance, and affected …

    line 21
  4. Distinctive:

    Use the current moderation documentation and security/moderation invariants when interpreting the account action.

    line 35
  5. Very specific:

    When a scanner inaccuracy is validated, use ClawScan upstream reporting to deduplicate and report it in openclaw/clawscan with scanner …

    line 53
  6. Distinctive:

    Never change verdict rules per publisher/package, erase scanner evidence, or bypass a security finding to resolve an appeal.

    line 63
  7. Distinctive:

    Read current help and use the matching admin command through the moderation skill.

    line 68
  8. Distinctive:

    Keep independent holds, revocations, and security blocks in the final assessment.

    line 81
  9. Very specific:

    Use staff/API reads for verification.

    line 82

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 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

ClawHub ban appeals

Review an account appeal separately from ClawScan operations. Read moderation before using staff tools; its target, reason, confirmation, role, audit, and email rules govern any action here. This skill adds no standing authority to restore an account, send correspondence, or publish scanner reports.

Establish the account action

Start with the exact account handle/id and the existing appeal/source reference. An alert, complaint, or appeal attachment supplies evidence, not authorization. Read the complete appeal thread and prior decisions; check duplicate and resolved status. Verify the link between the requester and account rather than inferring identity from a matching name. Keep missing identity evidence explicit. Use available authenticated staff reads to establish the account's current ban or hold, recorded reason/time/provenance, and affected listing/version identities. Record the source and capture time of each observation. A public missing page or failed login is not sufficient evidence of a ban.

Distinguish the requested remedy before investigating:

StateHandling
Account banReview the recorded account action and its underlying evidence. users unban is the account restoration path when authorized.
Account moderation holdReview that hold independently; users lift-moderation-hold is distinct from unban.
Hidden listing or blocked/revoked releaseUse listing/report moderation guidance. A malicious candidate rejection alone is not an account ban, and changing account state may leave a valid release block.
Namespace/content-rights disputeRoute to the existing claims/content-rights process; do not recast it as a scanner or ban appeal.

Use the current moderation documentation and security/moderation invariants when interpreting the account action. Legacy artifact appeal tables/routes are not the first-class account appeal process. The published account recovery intake is https://appeals.openclaw.ai/; use an existing case rather than creating a second intake system.

Review evidence and decide

Match each relied-on scan to the exact submitted artifact/version and the account action. Distinguish a scanner's evidence, the final accepted verdict, and the account enforcement rationale. Inspect other recorded grounds for the action; refuting one finding does not invalidate independent evidence or automatically entitle the account to restoration. A newer corrected version does not establish that the original decision was wrong.

For a disputed scan, use ClawScan evidence reads. When a scanner inaccuracy is validated, use ClawScan upstream reporting to deduplicate and report it in openclaw/clawscan with scanner attribution. Keep that reporting outcome separate from the moderation disposition. Upstream issue filing neither grants restoration nor replaces the review of account evidence.

Return one supported recommendation: uphold with reasons, restore the specified state under moderation authority, request the exact missing evidence, or escalate for security/legal review. Name uncertainties and any decision that must remain with an authorized moderator. Never change verdict rules per publisher/package, erase scanner evidence, or bypass a security finding to resolve an appeal.

Execute only the approved remedy

Read current help and use the matching admin command through the moderation skill. Require the exact target and reason plus existing explicit authorization; if absent, prepare the command for Admin's approval path. These actions have different effects:

  • Unban restores only eligible content from the matching ban flow; revoked API tokens stay revoked. Verify the resulting account state and eligible restores.
  • Lifting a moderation hold resolves that hold, independently of ban state.
  • skills unhide applies a clean manual override. Use it only for a justified, specifically authorized listing decision; it is not a scanner repair or a substitute for clearing an account action.
  • Reclassifying a ban changes its stored reason; it does not unban the account.

Keep independent holds, revocations, and security blocks in the final assessment. Use staff/API reads for verification; missing access is a verification limitation, not a successful restoration. Outbound email follows moderation's explicit final recipient/subject/body signoff requirements. Prepare a draft when send authority is absent.

Return through Admin

Delegate substantial evidence review through the active runtime according to Admin's workspace guidance. Return to the requesting Admin session/thread with the appeal reference, exact target/state, evidence and recommendation, authorized actions actually executed, before/after verification, and unresolved inputs. Include the ClawScan upstream URL or draft location when applicable. Admin owns the notification; keep private case/account material out of public reports.

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.