Ask the Oracle for a fast, cheap answer: the skill spawns a bare headless Fable session for ~1.3k tokens of overhead, and the oracle agent is a fallback subagent for when the headless call can't run (auth, sandbox), at ~6k tokens.
A Claude Code plugin marketplace with one plugin: oracle, a fast, cheap way to get a one-shot answer without derailing your main session's context or train of thought.
claude CLI binary on PATH — the skill shells out to a fresh headless claude -p session. If you're running Claude Code at all, you already have this.fable model. Your account needs access to it.ANTHROPIC_API_KEY is set in your environment, the skill runs headless with --bare (API rates, off your Claude Code quota). Otherwise it runs --safe-mode (OAuth, counts against your normal plan quota) — no extra cost beyond a small, cheap request.WebSearch. If a question needs a file's contents, the calling session reads the file first and pastes the relevant lines in.Inside a Claude Code session:
/plugin marketplace add benjaminjackson/oracle
/plugin install oracle@oracle
(The repo lives at benjaminjackson/oracle; oracle is the marketplace's internal name, used only in the plugin@marketplace install id above.)
Confirm it's installed: run /plugin (Installed tab) or /plugin list.
One skill, one agent, same job — answer a question as concisely as possible — at two different price points.
The default path. Writes your question to a scratch file, then shells out to a bare headless Fable session (claude -p --safe-mode or --bare, whichever fits your setup) with a stripped system prompt. No agent scaffolding, no tool loadout beyond WebSearch — about 1.3k tokens of overhead per question.
/oracle <question>macOS ships bash 3.2 by default, and a heredoc nested inside $(...) breaks the instant the body has a contraction (don't, it's) — a real parser bug, not a wording issue. The skill sidesteps it entirely: write the question to a plain text file with the Write tool (no shell parsing, so quotes and backticks are inert), then read it back with a bare $(cat file).
Fallback only, for when the headless call can't run — no claude CLI on PATH, auth broken, sandboxed environment with no shell access. Same brief (shortest correct answer, lead with the word that matters), run as a proper subagent instead of a headless process: Read, Grep, Glob, WebSearch, WebFetch, ~6k tokens per question.
The skill falls back to this agent automatically if the headless command fails, and says so when it does.
.claude-plugin/marketplace.json # marketplace manifest (this repo)
oracle/.claude-plugin/plugin.json # oracle plugin manifest
oracle/agents/oracle.md # fallback agent
oracle/skills/oracle/SKILL.md # headless-Fable skill
Benjamin Jackson (@benjaminjackson)
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
CRUD operations for Reclaim.ai tasks including create, update, list, complete, and delete with confirmation workflow
Ruby LSP integration with proactive SessionStart hook that guides Claude on when and how to use the 9 LSP operations for semantic code intelligence
Technical writing that respects the reader's time: the write skill drafts and updates READMEs, install guides, and dev docs, verifying every path, command, and flag against the actual code before describing it, refusing to document a UI flow it hasn't watched run, and cutting the reassurance, closing flourishes, and leftover conversation that bloat docs without telling the reader anything.
Structured reasoning skills for Claude Code: atom-of-thought decomposes complex multi-part problems into independent and dependent sub-questions before answering; confess produces an honest audit of just-completed work, surfacing confidence levels, shortcuts, and unverified assumptions; postmortem turns a finished session into a root-cause analysis, a minimum-round-trip replay, and proposed fixes.
Prose editing built on Deirdre McCloskey's Economical Writing: the editor agent critiques a draft line by line and refuses to rewrite it, the critique skill runs a document through editor and reports the findings with no edits made, and the tighten skill runs the same process but applies every safe and structural edit automatically, no approval step.
npx claudepluginhub benjaminjackson/oracle --plugin oracleComprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Harness-native ECC operator layer - 67 agents, 278 skills, 94 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Binary reverse engineering, malware analysis, firmware security, and software protection research for authorized security research, CTF competitions, and defensive security
Intelligent draw.io diagramming plugin with AI-powered diagram generation, multi-platform embedding (GitHub, Confluence, Azure DevOps, Notion, Teams, Harness), conditional formatting, live data binding, and MCP server integration for programmatic diagram creation and management.
Complete creative writing suite with 10 specialized agents covering the full writing process: research gathering, character development, story architecture, world-building, dialogue coaching, editing/review, outlining, content strategy, believability auditing, and prose style/voice analysis. Includes genre-specific guides, templates, and quality checklists.
Comprehensive .NET development skills for modern C#, ASP.NET, MAUI, Blazor, Aspire, EF Core, Native AOT, testing, security, performance optimization, CI/CD, and cloud-native applications