Subagent dispatch guidance for the Agent tool — whether to delegate or work inline, the four-part delegation contract, model and effort selection, parallel fan-out sizing, fork vs fresh subagent, and dispatching independent verifier agents. Use before any Agent tool call: spawning subagents, fanning out work across…
Controls a live Chrome browser through a persistent puppeteer-core MCP connection — click, type, hover, drag, fill forms; navigate pages; take screenshots and export PDFs; read the DOM and accessibility tree; run Lighthouse audits and Core Web Vitals traces; intercept, record, stub, or block network requests; export…
Decomposes user intent through structured brainstorming before acting on ambiguous or underspecified requests — classifies the ambiguity type, generates competing interpretations, and selects the clarifying question that removes the most uncertainty. Use when requirements are unclear, a request could have multiple…
Design and review prompts Claude-first — system prompts, reusable agent definitions, long or novel dispatch briefs, pipeline stages, and prompts users will run elsewhere. Two modes: design and adversarial review. Covers Claude model behavior and migration, de-prompting, few-shot design, output-schema ordering, context…
Captures and summarizes screenshots from the local desktop or a named window using the bundled capture script. Use when the user asks for a screenshot, screen capture, active-window capture, visual inspection of the current screen, or named-window screenshot. Not for browser-page screenshots when browser MCP tools are…
Create, evaluate, and review Claude Code skills and reusable agent definition files (structure, frontmatter, evals). Covers the skill-vs-subagent-vs-hook decision, SKILL.md authoring and frontmatter, trigger-description optimization, baseline-first evals with pressure testing via the skill-eval MCP tools, validation…
You are a fresh-context analyst for skill eval results. You were not involved in producing the skill or the runs and have no stake in the outcome. Work only from the files whose paths appear in your dispatch message.
Runs parallel multi-angle web research at four depth modes (scan, brief, breadth, deep): a planner grounds the query in local context and plans search dimensions, parallel search agents pull verbatim extracts with source URLs, one dimension always hunts counter-evidence, and synthesis is grounded back in the local…
Drafts, edits, and reviews prose to remove the AI tells a model cannot see in its own writing — measured detection signals, not intuition: structural rhythm first (sentence-length variance, information density, discourse order), a reader-job type router second (narrative, expository, persuasive, marketing, expressive…
Guides distributed system design using a structured 4-step framework. CHECKER mode audits existing designs. APPLIER mode walks through design step-by-step. Use when: system design, architecture review, scaling, capacity planning.
Guides distributed system design using a structured 4-step framework: scope, high-level design, deep dive, wrap-up. Two modes: CHECKER audits existing designs for missing components, wrong tradeoffs, and anti-patterns. APPLIER walks through designing systems step-by-step with clarifying questions, component selection…
Convert markdown to rich text styled for any platform — Slack, Teams, Discord, Notion, Word, Outlook, Gmail, Confluence, Jira, PowerPoint — and drop it on the clipboard ready to paste.
Set up penman's runtime after installing the plugin. Checks for bun (penman's runtime) and helps install it if missing. Run once after /plugin install penman@rtd.
Convert markdown to platform-styled rich text and copy it to the macOS clipboard, ready to paste with formatting intact into Slack, Teams, Discord, Notion, Word, Outlook, Gmail, Confluence, Jira, PowerPoint, Google Docs, Google Slides, or Craft — or copy a multi-line shell command as plain text for the terminal. In…
Read a past Claude Code session in this herdr space and get a clear catch-up summary, plus questions to resume it. Point it at an agent/space name and/or a session id, or nothing for the previous session. Never revives; reading only.
Finds, previews, and reads a past Claude Code session recorded on disk for a project or workspace ("space") — reading only (search, outline, verbatim turns), never reviving. Use when the user wants to find or look up a lost or previous Claude Code session or agent; asks what past Claude Code sessions exist for a…