expense-capture
Skill617 starsChanged 10 months agoMITView on GitHub
Capture a receipt or purchase as a structured, source-linked local expense record without guessing fields
We split this skill into the rules it gives an agent
8 rules. About 0 are common, 6 are distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Identify the source: Use details pasted by the user, an exact file path they provide, or a targeted email search they request.
- Distinctive: line 20
Mark unknowns: Never infer currency from locale, use the email date as the transaction date, expand a partial card number, or invent a tax …
- Very specific: line 21
Check likely duplicates: Use recall for records with the same merchant, date, and amount.
- Very specific: line 23
Store with provenance: Use remember for a concise episodic record, including the source kind and reference.
- Distinctive: line 24
Offer the next action: If reimbursement or follow-up is needed, propose a local, Todoist, or Asana task.
- Distinctive: line 28
Never claim an expense was filed, reimbursed, categorized by accounting, or synced externally unless a real tool reports that result.
- Distinctive: line 29
Never store full card or bank numbers, authentication data, or irrelevant personal details.
- Distinctive: line 30
If a file format cannot be read, preserve the partial structured record and ask for pasted text or a supported export.
≈ “Common” is an estimate: at least 1 in 5 of the 308,232 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
Expense Capture
Turn purchase details into a consistent local record. This skill records information in Hexis; it does not claim to submit an expense to an accounting or reimbursement system.
Workflow
- Identify the source. Use details pasted by the user, an exact file path they provide, or a targeted email search they request. Do not search broad mail history or filesystem locations to hunt for receipts.
- Extract without invention. Capture merchant, transaction date, amount, currency, tax or tip when explicit, category, payment method suffix when present, business purpose, reimbursable status, project or client, and a source reference.
- Mark unknowns. Never infer currency from locale, use the email date as the transaction date, expand a partial card number, or invent a tax split. Ask only for missing fields that materially affect the user's purpose.
- Check likely duplicates. Use
recallfor records with the same merchant, date, and amount. If one is similar, show it and ask whether to reuse, correct, or create a separate record. - Confirm the record. Present the normalized fields. An explicit request such as "capture this receipt" authorizes saving that shown record; an exploratory request does not.
- Store with provenance. Use
rememberfor a concise episodic record, including the source kind and reference. Avoid raw receipt bodies and unnecessary personal or payment data. - Offer the next action. If reimbursement or follow-up is needed, propose a local, Todoist, or Asana task. Create it only after the user chooses the task system and exact content.
Control and Failure Boundaries
- Never claim an expense was filed, reimbursed, categorized by accounting, or synced externally unless a real tool reports that result.
- Never store full card or bank numbers, authentication data, or irrelevant personal details.
- If a file format cannot be read, preserve the partial structured record and ask for pasted text or a supported export. Say what failed and what the user can do next.
Other versions1 project uses a version of this.
Word for word · 1
- QuixiAI/Hexis★ 617100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
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.
In this repository42 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- audio-analysisSkill
- calendarSkill
- code-executionSkill
- connector-action-authorizationSkill
- core-memorySkill
- cost-reportSkill
- councilSkill
- crm-lookupSkill
- daily-briefingSkill
- delegationSkill
- email-digestSkill
- github-issuesSkill
- gmail-actionsSkill
- gmail-connector-setupSkill
- home-assistantSkill
- host-nodeSkill
- humanizerSkill
- image-genSkill
- inbox-triageSkill
- integration-connector-setupSkill
- knowledge-ingestSkill
- meeting-follow-upSkill
- meeting-prepSkill
- memory-exchangeSkill
- notionSkill
- operationsSkill
- outreachSkill
- personal-notesSkill
- researchSkill
- self-inspectionSkill
- self-reflectionSkill
- skill-authoringSkill
- speechSkill
- spotifySkill
- travel-prepSkill
- trelloSkill
- twitter-x-actionsSkill
- twitter-x-connector-setupSkill
- weatherSkill
- weekly-reviewSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
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.

