Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Authors
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

ProTermsPrivacyRefunds
Back to skills

Gibson Resources

ASecurity

Discover which AI runners and infrastructure are actually available right now (Grok/Codex/Claude/Hermes CLIs, Devin API, GitHub auth, bot identities, satellite machines, MCPs) and produce the cost-optimized routing table the Gibson loop will use. Second stage of the /gibson pipeline; also standalone ('what runners do I have', 'fleet inventory', 'check the fleet').

26 stars
0 votes
0 copies
0 views
Added 9/28/2026
ai-agentsrustgogitapi

Works with

claude codecliapimcp

Security Analysis

A100/100

Scanned 9/28/2026

Install to Claude Code

$npx -y skills add The-AIE/the-gibson --skill gibson-resources --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Gibson Resources?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Gibson Resources
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/the-aie-gibson-resources/badge)](https://www.skillsdirectory.com/skills/the-aie-gibson-resources)

More formats (shields.io, HTML) on the badges page.

Files
SKILL.md
---
name: gibson-resources
description: "Discover which AI runners and infrastructure are actually available right now (Grok/Codex/Claude/Hermes CLIs, Devin API, GitHub auth, bot identities, satellite machines, MCPs) and produce the cost-optimized routing table the Gibson loop will use. Second stage of the /gibson pipeline; also standalone ('what runners do I have', 'fleet inventory', 'check the fleet')."
---

# gibson-resources — inventory before dispatch

Probe, don't assume. Every prior session's notes about what works may be stale —
tokens expire, CLIs update, machines go offline. Verify each resource this run.

**A `--version` check proves the CLI is installed, NOT that it's authenticated
or has quota.** Where a cheap auth check exists, run it and report auth state
separately from install state — an installed-but-unauthenticated runner is NOT
"alive" for routing. Where no cheap non-spending auth check exists (some CLIs
only reveal auth by doing billable work), say so and mark it "install-confirmed,
auth-unverified" rather than implying it's ready.

## The probe list

| Resource | How to verify (cheap, non-destructive) |
|---|---|
| **Grok CLI** | `~/.grok/bin/grok --version` (install). Auth: only confirmable by a real call — mark auth-unverified unless a prior lane succeeded this session. Flat-rate pool — the volume workhorse. |
| **Codex CLI** | `codex --version` (install); confirm it runs inside a trusted git dir. Auth surfaces on first real call — mark auth-unverified until one succeeds. Reviews need `-s read-only`. |
| **Claude Code** | you're running in it — note model + whether `claude` CLI headless is available for dispatched lanes. |
| **Hermes** | `hermes --version` if present. |
| **Devin** | `DEVIN_API_KEY` set? `scripts/devin-supervisor.sh --help` from the Gibson clone. **NEVER run `ensure` as a probe — if no session exists it CREATES one, which bills ACUs (real money). Spend needs the owner's explicit OK first (docs/14), a cap is not approval.** Report key-present/absent + `DEVIN_MAX_ACU` state only. |
| **GitHub** | `gh auth status`; which account; GraphQL vs REST budget state. Bot identities: check for GitHub App creds (e.g. `~/.claude/secrets/*.pem` + token-mint helpers) — distinct pusher/attester identities matter for review-gate integrity. |
| **Satellite machines** | ssh reachability of known satellites (e.g. `ssh -o BatchMode=yes -o ConnectTimeout=5 <host> 'echo ok'`). Report auth state honestly (a reachable box with dead GitHub auth is relay-only). |
| **MCPs** | note connected MCP servers relevant to the target (DB, Vercel, etc.). |

## Output: the routing table

Emit a table the `gibson-run` skill consumes directly, applying docs/15's
standing order (flat-rate absorbs volume; metered buys judgment):

- **implementer.default** → Grok (if alive), else Codex
- **implementer.second / reviewer.primary** → Codex (Grok reviews only when
  Codex is the author — Law 5 cross-vendor)
- **judgment + capability-gated work** → Claude: specs, escalation after N
  failures, adversarial verdicts, merge-bar calls — plus feature work whose
  capability bar the flat-rate pools genuinely can't clear (docs/15 routes by
  capability x marginal cost). Spend sparingly, reserve cap headroom.
- **merge_captain** → Devin IF key + supervisor wiring verified. ELSE fail
  closed in this strict order: (1) cross-vendor review remains MANDATORY while
  any second vendor is alive — Devin being unwired never relaxes Law 5; (2)
  only if NO other vendor is reachable at all, docs/11's fresh-context
  adversarial pass (different session, same platform) substitutes, flagged in
  the digest as degraded mode; (3) Tier C parks at the human gate in every
  mode. The routing table must say plainly that Devin is unwired and exactly
  what wiring it needs.
- Per-resource: verified-at timestamp, failure notes, cost tier.

Also report gaps as Ask Contract items when only the owner can fix them
(expired auth, missing API key, unprovisioned bot identity) — one compact list,
not a drip of interruptions.

Attribution

The-AIEThe-AIE
View sourceMore from The-AIE →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Ultra-compressed communication mode that cuts output tokens while keeping technical accuracy. Levels: lite, full, ultra and the wenyan variants. Use for /caveman, "caveman mode", "talk like caveman", "be brief" or "less tokens".

1074701 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

695601 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3351 votes

catchup

Recovers the conversation and failed tool calls of a previous Codex, Claude Code, Antigravity, Cline, Copilot CLI, Cursor, DeepSeek Harness, Kimi, OpenCode, Pi Agent, or ZCode session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", asks to recover/summarize a previous session before continuing, or asks to diagnose or report a catchup failure. Do NOT use for the current conversation, git history, or any non-agent log.

691 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →