agentleFS
NVIDIAKnown publisherNVIDIA/nemoclaw-community/examples/recipes/nvidia/payment-ops-hermes/skills/release-packet/SKILL.md

release-packet

Skill142 starsChanged 3 months agoApache-2.0View on GitHub

Prepare a human-approval release packet for a screened payment. Does not release funds — release requires a human checker.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 2 are distinctive, and 3 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.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use this skill after payment-screening returns CLEARED_FOR_REVIEW, when an operator wants the payment prepared for a human approver.

    line 8
  2. Distinctive:

    Do not attempt it. If you try to POST to the payment rail, the sandbox drops the connection (deny-by-default).

    line 40
  3. Very specific:

    Never set status to anything implying the payment was sent.

    line 46
  4. Distinctive:

    Never claim funds moved.

    line 47
  5. Very specific:

    Do not prepare a packet for a payment that screened as HOLD.

    line 48

≈ “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-packet

Use this skill after payment-screening returns CLEARED_FOR_REVIEW, when an operator wants the payment prepared for a human approver.

This skill prepares a packet marked PENDING_HUMAN_APPROVAL. It does not contact the payment rail. FinGuard cannot reach payments-rail.internal — that egress is denied by the sandbox policy. Only a human approver on the host can release.

Procedure

1. Re-screen, then prepare the packet

/usr/bin/python3 /sandbox/.hermes/skills/release-packet/scripts/prepare_release_packet.py \
  --queue /sandbox/.hermes/skills/payment-screening/data/payment-queue.json \
  --sanctions /sandbox/.hermes/skills/payment-screening/data/ofac-sdn-fixture.json \
  --id WIRE-1007 \
  --maker "$USER"

The packet includes the payment, the screening result, the maker identity, and status: PENDING_HUMAN_APPROVAL.

2. Hand off to a human checker

Tell the operator, explicitly, that release is a human action:

"Release packet prepared for WIRE-1007, status PENDING_HUMAN_APPROVAL. I cannot release this — a human approver must run the release on the host."

3. If asked to release anyway

Do not attempt it. If you try to POST to the payment rail, the sandbox drops the connection (deny-by-default). That is by design: maker-checker is enforced by the platform. Restate that a human approver is required.

Pitfalls

  • Never set status to anything implying the payment was sent.
  • Never claim funds moved.
  • Do not prepare a packet for a payment that screened as HOLD.

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.