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.

redstrike-excessive-agency - Skill

Name: redstrike-excessive-agency
Author: 0xnookie

Imported from 0xnookie/redstrike-ai (.claude/skills/redstrike-excessive-agency/SKILL.md). Install upstream with npx skills add 0xnookie/redstrike-ai --skill redstrike-excessive-agency. Copyright stays with the author.

Run the Redstrike excessive-agency methodology (canonical spec: skills/excessive-agency/). OWASP LLM06.

Scope + approval: read skills/_contract/roe-gate.md + the engagement scope.yaml. The LLM performs REAL actions via tools — get approval + kill switch; confirm with BENIGN effects only (OOB callback / no-op marker action); never destructive (no real emails/payments/data changes). SSRF via tools stays in scope (metadata → ssrf-cloud).
Execute: follow skills/excessive-agency/SKILL.md; consume <engagement>/llm-context/llm_context.json (tools); work skills/excessive-agency/checklist.md; use skills/llm-context/refs/. Write outputs to <engagement>/excessive-agency/. Downstream tool-arg injection → hand to ssrf-*/sqli-*/command-injection.
Coverage: copy your filled checklist into <engagement>/excessive-agency/checklist.md.
Mechanical gate: python skills/_contract/verify_completion.py --skill skills/excessive-agency --out <engagement>/excessive-agency → fix + re-run until exit 0.
Reviewer: launch the excessive-agency-verify subagent on <engagement>/excessive-agency. Only accepted findings go to reporting.

Install redstrike-excessive-agency 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": "0xnookie-redstrike-ai-redstrike-excessive-agency", "kind": "skill", "name": "redstrike-excessive-agency", "description": "Run the Redstrike excessive-agency hunt (OWASP LLM06) — abuse the LLM's tools (param injection, SSRF via tools, tool chaining) to drive a benign unauthorized action — then gate and hand off to the excessive-agency-verify reviewer agent.", "publisher": "0xnookie", "version": "1.0.0", "capabilities": { "domains": [ "coding" ], "tags": [ "skill-md", "github" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "Run the Redstrike excessive-agency hunt (OWASP LLM06) — abuse the LLM's tools (param injection, SSRF via tools, tool chaining) to drive a benign unauthorized action — then gate and hand off to the excessive-agency-verify reviewer agent." ], "primary": false, "metadata": { "source": { "provider": "github", "repository": "https://github.com/0xnookie/redstrike-ai", "path": ".claude/skills/redstrike-excessive-agency/SKILL.md", "ref": "5f716e2483d0279c3fdfc9aac13a35392b53cf29", "url": "https://github.com/0xnookie/redstrike-ai/blob/5f716e2483d0279c3fdfc9aac13a35392b53cf29/.claude/skills/redstrike-excessive-agency/SKILL.md", "key": "0xnookie/redstrike-ai/.claude/skills/redstrike-excessive-agency/SKILL.md" } }, "instructions": "Run the Redstrike **excessive-agency** methodology (canonical spec: `skills/excessive-agency/`). OWASP LLM06.\n\n1. **Scope + approval:** read `skills/_contract/roe-gate.md` + the engagement `scope.yaml`. The LLM performs REAL actions via tools — **get approval** + kill switch; confirm with BENIGN effects only (OOB callback / no-op marker action); **never** destructive (no real emails/payments/data changes). SSRF via tools stays in scope (metadata → `ssrf-cloud`).\n2. **Execute:** follow `skills/excessive-agency/SKILL.md`; consume `<engagement>/llm-context/llm_context.json` (`tools`); work `skill", "cost": { "context_tokens": 299 } }

redstrike-excessive-agency

About

Use it

Install redstrike-excessive-agency 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

Manifest

Reviews