Agent skills for Devin Desktop

Runs a long, structured multi-step reasoning chain over the sequential-thinking MCP server, with branching, revision, and explicit…

Use for on-demand deep exploration of an existing codebase - analyzes architecture, data flow, domain model, API surface, dependencies…

Deep autonomous diagnosis of a failed flow run.

Optimize deep learning models using Adam, SGD, and learning rate scheduling

Procedure for answering a question with a thorough, cross-checked, cited web-research report using the Keenable search + fetch tools.

Use DeepAPI for all regular web searches (POST /v1/search/web) and for any deep research you perform (POST /v1/research/deep), instead of…

Export raw OpenTelemetry traces from an AI application to Confident AI's Observatory.

Configure Deepgram CI/CD integration for automated testing and deployment.

Implement production pre-recorded speech-to-text with Deepgram.

Implement real-time streaming transcription with Deepgram WebSocket.

Collect Deepgram debug evidence for support and troubleshooting.

Deploy Deepgram integrations to production environments.

Execute Deepgram incident response procedures for production issues.

Use for evidence-first reading of articles, books, PDFs, web pages, or document sets, with knowledge maps and Feynman checks.

Reference documentation for DeepWork Reviews — automated code review rules using .deepreview configs and DeepSchema-generated rules

Analyze a RAG task and produce guidance for choosing exactly one available Agentic RAG Skill using evidence dependencies, retrieval…

Deferred shading/lighting skill for G-buffer design and decals.

Generate PRODUCT_SPEC.md through guided Q&A.

Writes work items to a business-analyst standard, with testable acceptance criteria, so complete an implementer needs to ask nothing.

Extract clean markdown content from web pages using Defuddle CLI, removing clutter and navigation to save tokens.

How to delegate all AI work to the agent chat.