worklog-audio
Skill4 starsChanged 4 months agoMITView on GitHub
End-of-complex-work habit — narrate what was just done as an audio note via GCP TTS (Gemini-TTS, latest model in Model Garden) and deliver the voice file. Trigger at the end of any substantial working session ("audio note", "voice summary", automatic habit after complex work), or when the user asks to "explain what you did" as audio.
In this repository16 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- llms-full.txtllms.txt
- llms.txtllms.txt
- agents-syncSkill
- call-agentSkill
- captureSkill
- context-restoreSkill
- context-saveSkill
- curatorSkill
- learnSkill
- onboardSkill
- recallSkill
- reviewSkill
- skillifySkill
- wiki-doctorSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 12
The habit: every time a complex piece of work finishes, produce an audio note through OUR GCP that explains the work end-to-end in layman …
- Distinctive: line 32
Default voice Charon (best clear-narration voice of the Gemini-TTS set.
≈ “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.
Files
worklog-audio — spoken work-log via GCP
The habit: every time a complex piece of work finishes, produce an audio note through OUR GCP that explains the work end-to-end in layman terms, and hand over the voice file.
Steps
- Write the narration to
artifacts/<date>-<slug>/worklog-narration.md. Plain spoken prose, no jargon left unexplained, 400–900 words. It MUST cover, in order:- What was done — end to end, in plain language a non-engineer follows.
- How it was done — the technical approach, translated to lay terms (analogies over acronyms; expand every acronym once).
- What it means for us / the business — impact on qcobrain/Quarktex, concretely.
- How well it went — honest self-assessment incl. what failed or is unverified.
- What's next — the next actions, and any decision the user owns. Ground every factual claim in the session or cited notes — no invented numbers.
- Pick the model by probing, not memory. Latest verified:
gemini-2.5-pro-tts(2026-08-02 — "3.6" does not exist; Model Garden list API returns nothing for TTS, so probetext:synthesizewith candidate names and use the newest that answers OK). - Synthesize:
bash skills/worklog-audio/narrate.sh <narration.md> <same-dir>/worklog.mp3Default voice Charon (best clear-narration voice of the Gemini-TTS set; alternatives: Kore, Aoede, Enceladus). Voice is arg 3. - Verify before claiming done:
afinfo <file>(oravmediainfo) — check the duration is plausible for the word count (~150 wpm), not that the file merely exists. - Deliver: give the absolute file path (and
open <file>command) in the final reply, and mention it in the session'slog/build-log.mdentry.
Notes
- Runnable half lives in
skills/worklog-audio/narrate.sh(model-agnostic bash+python, chunks long text, concatenates MP3). - Costs pennies per note (Gemini-TTS is ~$16/1M chars; a 5-min note ≈ 4k chars ≈ $0.06).
- This is a DELIVERABLE, so it belongs in
artifacts/, never the scratchpad.
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.

