Messy pasted text — lists, receipts, leads, citations — into a clean, typed table. Outputs Markdown + CSV, and flags ambiguous rows instead of guessing.
Lean agent skills.
Install once.
Reach for one when you need it.
A small, curated marketplace from Aonami Labs. Each skill does one thing, carries no heavy dependencies, and loads only when it's relevant.
Two commands, then mention it.
Add the marketplace, install the pack(s) you want, and call any skill by name. No build step, no config.
Not on Claude Code? SKILL.md is an open standard — point Codex CLI, Cursor, Gemini CLI, or Copilot at plugins/**/skills.
Thirteen skills. Two packs.
Summarises a contract or policy, lists obligations, deadlines & amounts, and flags auto-renewal / liability clauses — each citing the clause it came from.
Turns raw notes or a transcript into a summary, a decisions list, and an owner · task · due action table. Copy-ready checklist included.
Realistic synthetic test data for any described schema — users, orders, events — as JSON/CSV, with referential integrity and a seed for reproducibility.
Profile a CSV or dataset — per-column type, nulls, distributions, top values — and flag data-quality issues like mixed types, outliers, and lost leading zeros. Describes; never modifies.
Turn commits or merged PRs into a grouped changelog — Added / Changed / Fixed / … — following Keep a Changelog, with a suggested SemVer bump. Won't invent changes.
Draft a new email or reply from a brief or a pasted thread, in your tone and length. Returns subject options, a TL;DR, and a ready-to-send body. Drafts; doesn't send.
Turn a git diff or PR into a review-ready summary grouped by area, with risk callouts — breaking changes, migrations, missing tests — and a suggested PR title + description.
Generate mock API responses from an endpoint or schema — the success payload plus error and edge states (4xx/5xx, empty, paginated), seeded and consistent. Doesn't call the API.
Decode an AA/DEPA consent artefact — JSON, full wrapper, or signed JWS — into plain English: who wants the data, what, why, how often, how long, and retention. Flags broad scope & indefinite retention.
ReBIT deposit statement → monthly cashflow, detected income & regularity, recurring obligations, FOIR, surplus, risk flags, and a transparent cashflow score that shows the math.
Schema-valid ReBIT deposit test data (Profile / Summary / Transactions) in XML + JSON, with reconciling balances — from a profile like salaried, gig, or MSME.
Parse a upi://pay link or QR intent into "you're about to pay X to Y". Flags an editable or missing amount and a missing payee name.
Our own tools, released.
Aonami Labs builds internal tools and releases them for developers to use, fork, and improve. These skills are the small, sharp pieces of our own workflow — drop them into your agent and get value in seconds, not set up a framework.
The India Stack skills are built on India's open digital public infrastructure — Account Aggregator / DEPA and UPI. Those rails belong to the ecosystem (RBI/ReBIT, NPCI, Sahamati and the wider community); we contribute developer tooling on top, and welcome collaboration.
A note on honesty: skills that decode do not verify. consent-decoder reads an artefact and tells you what it says — it does not check the signature's authenticity. The skills state what they don't do, on purpose.
One thing, done well.
Every skill in this marketplace clears the same four lines before it ships.
