agentleFS
ric03uecric03uec/clawrium/skills/hermes/release-watcher/SKILL.md

release-watcher

Skill67 starsChanged 2 months agoApache-2.0View on GitHub

Watch upstream *Claw releases and clawrium discussions; surface top 3 feature candidates to Devashish via Discord DM for approve/skip.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    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. Very specific:

    Wait up to 24 hours for a reply.

    line 78
  2. Very specific:

    Do not file issues without an explicit approve N from Devashish.

    line 98
  3. Very specific:

    Never apply the agent-ready label (PAT-blocked).

    line 99

≈ “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

release-watcher

Once per day, scan upstream *Claw releases and the ric03uec/clawrium GitHub Discussions for feature ideas that could land in clawrium. Surface the top 3 candidates to Devashish via Discord DM with an approve / skip pattern. On approval, file a GitHub issue.

Inputs

  • Releases (last 7 days, GitHub API):
    • openclaw/openclaw releases
    • zeroclaw/zeroclaw releases
    • nousresearch/hermes releases
    • Any additional supported *Claw repos listed in platform/registry/*/manifest.yaml (look up dynamically; don't hardcode).
  • gh api repos/ric03uec/clawrium/discussions — open discussions updated in the last 7 days, sorted by reaction count.
  • Memory key release-watcher:seen — release tag IDs already presented, so candidates don't repeat across days.

Scoring

Pick the 3 candidates that score highest on:

  1. User-facing: does this change something a clawrium user would directly touch (CLI, GUI, agent behavior, install flow)?
  2. Fits clawrium's surface: matches an existing primitive (agent type, integration, channel, skill) rather than introducing a new top-level concept.
  3. Cheap to land: rough fit in an afternoon's work, not a quarter.

Skip:

  • Pure internal refactors in the upstream project.
  • Features that depend on infrastructure clawrium does not have (custom kernels, GPU sharding, etc.).
  • Anything already covered by an open issue (check before suggesting).

Output — Discord DM

Send a single DM to user 740723459344302120 (Devashish) in the project guild. Slack is not used — the original issue proposed Slack DM for this flow; release-watcher uses Discord DM in the same guild instead.

Format:

Release watch — <YYYY-MM-DD> — 3 candidates

1. <project> <tag> — <one-line rationale>
   <link>

2. <project> <tag> — <one-line rationale>
   <link>

3. <project> <tag> — <one-line rationale>
   <link>

Reply: `approve 1 2` to file issues, `skip all`, or `skip 3` to drop
the third while keeping 1+2 in tomorrow's pool.

Wait up to 24 hours for a reply. If no reply by next run, treat as skip all and move on — do not nag.

On approval

For each approved candidate, create a GitHub issue on ric03uec/clawrium:

  • Title: <verb> <feature> from <project> <tag>
  • Body: link to the upstream release notes + the one-line rationale.
  • Labels: best-effort type:* and area:* from the candidate description. Do not apply agent-ready.
  • Assign no one — Devashish chooses the owner later.

Append the candidate's release tag to release-watcher:seen.

Hard constraints

  • One DM per day. If nothing crosses the bar, send a one-liner: "Release watch — quiet day, no candidates." Silence reads as broken.
  • Do not file issues without an explicit approve N from Devashish.
  • Never apply the agent-ready label (PAT-blocked).
  • Discussions are read-only here — do not post a comment in the discussion thread itself.

Anti-patterns

  • Presenting 5 candidates because they all "look promising." 3 is the cap; rank harder.
  • Re-presenting yesterday's skip items the next day. Use the memory key.
  • Filing an issue then immediately self-assigning a triage label. Triage is issue-triage's job; let that skill run on the new issue.

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.