Diagnose WebGL black or blank canvases: zero-sized canvas, lost context, missing draws, failed programs, incomplete FBOs, empty viewport/scissor, disabled color mask, or culled geometry. Use when the canvas stays black or empty despite an active WebGL app.
Apply the CMI Agent Operating System contract for substantive work: orient with explicit goal, scope and authorization boundaries; observe before editing; capture typed evidence; diagnose before symptom fixes; prioritize explicitly; implement the smallest coherent change; verify the original journey or contract…
Research and synthesize substantial live X evidence across one or multiple topics, labeled query angles, Top and Latest, relevant timelines, and post context. Use for open-ended research, comparisons, trend discovery, emerging accounts, current events, or technical discussion on X.
Set up, repair, or reconfigure DSH Codex Bridge for the current Git project. Use when the user installs the plugin, asks to configure DSH, add or switch a model, change the default Profile, preview model configuration, roll back a Profile change, or when delegation reports missing/invalid Bridge configuration.
Turn a captured Codex tool failure into reviewed .tape evidence and an offline regression test. Use when the user asks to inspect a failed run, fork captured evidence, save a regression, or prepare AgentTape evidence for CI.
A set of instructions for coordinating work between Codex and DSH through an existing tool connection. DSH is treated as a worker for bounded coding tasks, while Codex keeps responsibility for planning, technical decisions, supervision, and final review.
Recall earlier conversations from the user's other AI tools (Claude Code, Claude Desktop, Codex, Cursor, Gemini CLI, Antigravity, Devin). Use before asking about or assuming a project's history, and when the user refers to a prior decision, discussion or fix: "as we discussed", "the earlier one", "what we did with…
Audit and correct accessibility problems in rendered web interfaces using the wcag-accessibility MCP tools. Use when an agent creates, changes, reviews, or verifies HTML/CSS/UI code; when asked about WCAG, contrast, labels, semantics, keyboard access, focus, screen readers, responsive accessibility, or accessibility…
Inspect, validate, search, extract, copy-edit, native-open, or export local Hangul HWPX files on Windows, including table morphology and report paragraph hierarchy analysis, hanging-indent/style normalization, and interleave-page exclusion. Use when the user names an .hwpx file, asks to preserve Hangul layout, edit or…
Use for EVERY coding request when the user writes in their native language and wants to learn another language while building. Translates the prompt via Polyprompt, works in the target language, and surfaces vocabulary for spaced-repetition review.
Search for, select, and download books through the download-books MCP server. Use when a user asks an agent to fetch books into a local library folder, choose the best candidate from search results, or retry download candidates without exposing provider internals. Requires the download-books MCP tools.
Research mobile apps, broader-market competitors, positioning, public App Store reviews, onboarding, paywalls, recorded screens, user flows, App Store creatives, developers, and saved collections through the ScreensDesign MCP. Use when the user asks for mobile product or market research, app ideas, competitor…
Connect an existing agent (Hermes, Claude Code, Codex, any stdin/stdout CLI, an HTTP endpoint, or an MCP client) to Slack Agent Mesh, and operate correctly inside it — addressing agents by @name, delegating agent-to-agent, and respecting loop limits and untrusted-content rules. Use when wiring an agent into a shared…
Use Airlock before installing or executing packages, running risky shell commands, committing generated code, or declaring an agent task complete. Airlock detects hallucinated/slopsquatted dependencies, destructive commands, leaked secrets, and suspicious test changes.
★not rated 1 2mo agoA✓ AI review50 tokens
originalMIT
Set up, refresh and reconcile projects and runnable services with a self-hosted DevHub registry, then answer operational questions from its catalog. Use when the user asks to set up DevHub, connect tools or everything they can access, build their map, refresh DevHub, register, add, sync, reconcile, update, find, open…
A specialized guide for analyzing a custom virtual machine written in JavaScript, including its instruction codes and runtime behavior. A virtual machine here is a program that interprets its own small instruction language.
★not rated 1 todayA0 tokens
copy · 84%MIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: