How long does the integration take?

Change one base URL in an OpenAI-compatible client, or install the Python or TypeScript SDK; the quickstart routes a first request in a few minutes.

Will this lock us into another vendor?

No. The router is Apache-2.0 open source, your provider keys stay yours, and the API is OpenAI-compatible, so leaving is a change of base URL.

What happens when a provider fails?

The router falls back to the next model your policy allows and records the fallback in the trace, so an outage costs a retry, not an incident.

audit-verification - Skill

Name: audit-verification
Author: garbageek

Imported from garbageek/claude-plugins (skills/audit-verification/SKILL.md). Install upstream with npx skills add garbageek/claude-plugins --skill audit-verification. Copyright stays with the author.

Audit Verification

Read the runtime invocation contract first. Use the bundled Python CLI after resolving AUDIT_PLUGIN_ROOT and AUDIT_PROJECT_DIR as described there. A separately configured audit_* MCP server is an optional equivalent with an explicit absolute root.

Purpose: independently verify resolved tickets against the original acceptance criteria and current code state.

If this agent/session implemented the fix, hand verification to a fresh native subagent or another session with this skill, the project root, ticket IDs, fix revision, and original criteria. Do not treat switching the role name as independent verification. Without an independent reviewer, leave the ticket awaiting verification and report the handoff; do not write PASS.

Use the shared resolution proof guide to separate proposed checks, observed results, and verdicts, including whether corrected source is active in the relevant running instance. Keep the existing acceptance gates and user-authorized validation scope; do not generate tests or restart services implicitly.

First move from empty context

python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" doctor
python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" next --for verification --json

If the workspace is absent or no ticket is returned, report that state without initializing it. Do not create or resolve tickets. Diagnose malformed records rather than treating a failed queue command as an empty queue.

PASS flow

python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" show 001
# inspect code and evidence independently
python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" verify 001   --status PASS   --verified-commit def456   --criterion "AC1: pass - null-user payload returns typed validation error"   --evidence "Manual reproducer no longer fails"   --test "manual reproducer: pass"   --verdict "All acceptance criteria satisfied"   --as audit-verification   --json

Repeat --criterion once per single-line result using the original unique AC IDs. Evidence, the executed check/result, and the verdict must be meaningful in the active record; placeholder text or quoted examples do not satisfy PASS. Use PARTIAL or FAIL when evidence is incomplete or a criterion is not satisfied. Do not replace an existing verification file with the reference template.

Allowed actions

read ticket, verification record, code, and resolution evidence;
run independent checks/commands needed to validate the result;
set PASS, PARTIAL, FAIL, REGRESS, BLOCKED, WONTFIX, or INVALID through python3 "${AUDIT_PLUGIN_ROOT}/scripts/audit.py" --root "${AUDIT_PROJECT_DIR}" verify.

Forbidden actions

do not edit implementation code;
do not perform resolution work;
do not mark PASS without criterion evidence and a verified commit;
do not invent missing acceptance criteria.

Stop condition

Stop after writing the independent verdict and summarize the reason.

Install audit-verification and it becomes one more candidate the router can pick - when it fits.

A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.

{ "ocm": "1", "id": "garbageek-claude-plugins-audit-verification", "kind": "skill", "name": "audit-verification", "description": "Use when resolved audit tickets need independent verification. Default to the next READY_FOR_VERIFICATION ticket. Verification may write PASS/PARTIAL/FAIL with criterion evidence; it should not edit implementation code.", "publisher": "garbageek", "version": "1.0.0", "capabilities": { "domains": [ "coding", "customer_support" ], "tags": [ "skill-md", "audit", "verification", "validate", "verdict", "evidence", "github" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "Use when resolved audit tickets need independent verification. Default to the next READY_FOR_VERIFICATION ticket. Verification may write PASS/PARTIAL/FAIL with criterion evidence; it should not edit implementation code." ], "primary": false, "metadata": { "source": { "provider": "github", "repository": "https://github.com/garbageek/claude-plugins", "path": "skills/audit-verification/SKILL.md", "ref": "2aefa303113af347b915f5d982d5506bde239f16", "url": "https://github.com/garbageek/claude-plugins/blob/2aefa303113af347b915f5d982d5506bde239f16/skills/audit-verification/SKILL.md", "key": "garbageek/claude-plugins/skills/audit-verification/SKILL.md" } }, "instructions": "# Audit Verification\n\nRead the [runtime invocation contract](../../docs/audit/CONTRACT.md#1-canonical-runtime-invocation) first. Use the bundled Python CLI after resolving `AUDIT_PLUGIN_ROOT` and `AUDIT_PROJECT_DIR` as described there. A separately configured `audit_*` MCP server is an optional equivalent with an explicit absolute `root`.\n\nPurpose: independently verify resolved tickets against the original acceptance criteria and current code state.\n\nIf this agent/session implemented the fix, hand verification to a fresh native\nsubagent or another session with this skill, the project root, tic", "cost": { "context_tokens": 799 } }

audit-verification

About

Audit Verification

First move from empty context

PASS flow

Allowed actions

Forbidden actions

Stop condition

Install audit-verification and it becomes one more candidate the router can pick - when it fits.

What it costs to adopt - in engineering time.

One base URL

No vendor lock-in

Failover built in

Use it

Manifest

Reviews