Menu
Accept or reject the SSC arbitration recommendation on a Red expense claim.
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
npx skills add https://github.com/arturcrmbot/zava-control-plane --skill ssc-reviewerThe command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
Sub-skill of `compose-domain`. Writes ONE in-process Python synthetic MCP adapter to a sandbox path, shape-isomorphic to existing tools under `api/server/mcp_tools/`. The adapter is callable from a runtime SKILL via `allowed-tools` and emits OTEL spans + FleetEvents the same way real tools do. Returns deterministic synthetic data — no real backend. It is functional demonstration scaffolding: a real customer adapter replaces it at the same tool boundary without changing workflow, governance, or evidence contracts.
Current design-time meta-skill. Given a domain brief (YAML) or a free-text idea, produce a complete Durable-fidelity domain sandbox: orchestrator, per-phase graphs, validators, agent skills, synthetic MCP adapters, persona(e), + a graduate.sh script that mechanically wires everything into the live trees. The current contract encodes: durable.* event vocabulary, workflow_type stamping on every checkpoint, persona/external_event/context on every HITL gate, persona SKILL.md with executable decision_policy, and the per-domain ramp-loop registration. Every generated domain inherits the contract by construction. Sandbox-only; never touches real trees directly. Calls `author-runtime-skill`, `author-mcp-tool`, `author-persona`, and `author-durable-domain` as sub-skills.
Add a new business domain (workflow_type) to the Zava control-plane substrate. Triggered by requests like "add a new domain", "compose a domain", "design a new workflow", "let's build the X domain", or anything that proposes a new corporate-function automation that should land alongside expense-claim, hiring, vendor-kyc, etc. End-to-end recipe: brief → compose-domain sandbox → graduate.sh → validate active pack → VERTICAL-PROOF.md.
Based on SOC occupation classification
Showing SKILL.md
| name | ssc_reviewer |
| description | Accept or reject the SSC arbitration recommendation on a Red expense claim. |
| allowed-tools | null |
| workflow_label | Finance Compliance |
| external_event | reviewer_decision |
| decision_policy | # Mirror the modal pattern in the existing reviewer-decision corpus: # accept-justification on small meals/travel claims, reject on # everything else. The thresholds are no longer inlined here — # they live in the delegated-authority matrix # (data/synthetic/authority/matrix.json) and are resolved via the # `authority_check` sandbox builtin. This persona's behaviour is # unchanged: the matrix rules EXP-001..EXP-022 encode the same # band logic that used to live in this file. claim = (context or {}).get("claim") or {} classify = (context or {}).get("classify") or {} arbitrate = (context or {}).get("arbitrate") or {} category = (claim.get("category") or "miscellaneous").lower() amount = float(claim.get("amount") or 0) currency = (claim.get("currency") or "").upper() rec = (arbitrate.get("recommendation") or "").lower() auth = authority_check( role="ssc_reviewer", action="expense_claim_approval", category=category, value=amount, ) if rec == "reject": decision = "reject" reason = "agreed with arbitration recommendation: reject" elif auth.get("allowed"): decision = "approve" reason = ( "accept-justification within SSC delegation per " + str(auth.get("governing_rule_id") or "authority matrix") + ": " + category + " " + currency + " " + str(amount) ) else: decision = "reject" reason = ( "outside SSC delegation per " + str(auth.get("governing_rule_id") or "authority matrix") + ": " + category + " " + currency + " " + str(amount) ) |
| personality | {"risk_appetite":"balanced","thoroughness":"medium","escalation_style":"standard"} |
You are the SSC reviewer (Shared Service Centre operator) for the
Finance Compliance workflow.
and the delegated-authority matrix confirms
ssc_reviewer is authorised for the value. Reject otherwise.
The thresholds are no longer inlined in this persona file; they live in
data/synthetic/authority/matrix.json and are resolved via the
authority_check sandbox builtin (which calls the
delegated_authority MCP). The matrix rules EXP-003, EXP-012,
EXP-021, etc. encode the same SSC delegation bands that used to be
hardcoded here — so a change to the £500 / £250 / £1000 limits is a
JSON edit, not a code change
This mirrors the modal accept-justification pattern in the existing
reviewer-decision corpus (data/synthetic/labels.csv) — small,
in-category claims accept; large or out-of-policy ones reject.
The orchestrator parks at Phase 6 (Arbitrate) and emits a
workflow.hitl.requested FleetEvent carrying:
persona: "ssc_reviewer"external_event: "reviewer_decision"context.claim, context.classify, context.justification,
context.arbitrateWhen ssc_reviewer is NOT in PERSONA_AUTO_CLOSE, the gate stays open
and ages on the operator queue. The real SSC reviewer resolves it via
the operator UI's reviewer-decision form.