voice-input
Skill0 starsChanged 37 days agoMITView on GitHub
Read dictated (speech-to-text) requests the way the speaker meant them. Use when a request looks voice-typed, with misheard names, glued or split words, spoken numbers, odd punctuation, or a missing "not".
1 other project uses a version of this.
Adopted by
Most recently taken up by
- christos-antono/NPCTalktheir own version, first seen Oct 17, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- christos-antono/NPCTalk★ 231same 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
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 21
Recover the meaning yourself first: Use the active project, recent messages, the project list, and the glossary.
- Very specific: line 23
Don't quote the mistake and don't correct the user: If the recovery is not obvious, start your reply with one line: Understood as: ..., …
- Very specific: line 43
Keep a table of known mishearings in templates/glossary.md (or inline here) and extend it whenever a new one appears.
≈ “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
Voice input
The person you work with dictates most requests through speech-to-text (Wispr Flow, macOS dictation, Whisper, phone voice typing). Treat recognition errors as typos, not as intent.
What goes wrong in dictation
- Proper nouns: project names, brands, domains, file names, commands.
- English terms written phonetically in another script, or words split and glued wrongly ("get hub", "post gres").
- Numbers, amounts, IDs, email addresses, keys. These break most often.
- Punctuation: commas and periods in random places. Two requests may fuse into one sentence; one may split into two.
- Fragments: the start of a sentence is missing, or a stray word trails at the end.
- Negation: "don't" or "not" dropped or inserted. The most dangerous category.
How to work
- Recover the meaning yourself first. Use the active project, recent messages, the project list, and the glossary. A name that sounds like a known project almost always is that project.
- Don't ask because the form is messy. Grammar, word order, and odd endings are ignored silently.
- Don't quote the mistake and don't correct the user. If the recovery is not obvious, start your reply with one line:
Understood as: ..., then do the work. - Two equally likely readings: ask. One short question with the options, not a list of clarifications.
- Irreversible step on a heard value: verify against the source. Domain, amount, email, ID, table name: check the config, dashboard, or database before deploying, sending, paying, or deleting.
Always ask when
- The target is ambiguous: two similar-sounding projects, accounts, domains, or mailboxes.
- You can't tell whether a negation is there ("push" vs "don't push").
- It involves money, production, deletion, or sending something outside (email, messages, posts, submissions). The bar for asking is lower here. "Almost sure" doesn't cancel the cost of being wrong.
For everything else, a needless question caused by bad audio is friction, not caution. Act.
Example
Heard: hey deploy the pine cone checkout fix no wait dont push yet just commit it and bump to one point four oh
Reply starts: Understood as: commit the checkout fix in pinecone-shop, do not push, version 1.4.0.
Glossary
Keep a table of known mishearings in templates/glossary.md (or inline here) and extend it whenever a new one appears.
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.

