AI & ML

jscraik Bundle

Provides detailed answers, conceptual explanations, workflow guidance, and framework-based analysis about Edouard Maleix's talk "How AI-First Dev Teams Build Collective Intelligence — One Attributed Mistake at a Time." Use when the user asks about giving coding agents their own identity and signed commits, the diary/entry/pack/render workflow, turning agent mistakes into reusable team knowledge, evaluating knowledge packs for fidelity and usefulness, voluntary task picking by autonomous agents, the MoltNet open-source project, compound engineering, or applying his approach to make agent lessons compound across a team instead of evaporating into chat history.

jscraik Bundle

Explains Robert Overweg's One Brain, No Filtering talk and helps design safe knowledge-memory systems: context maps, retrieval rules, provenance labels, local knowledge-store structure, and review checkpoints. Use when the user asks about agent memory, unified knowledge bases, reducing context loss, or designing inspectable knowledge workflows.

jscraik Bundle

Use when the user asks about Amit Kushwaha's AI Native DevCon talk on benchmarking agent-era systems, measuring performance beyond single LLM calls, inference, workflow complexity, tool use, and real-world workloads.

jacob-balslev Skill

This skill provides decision frameworks for choosing the optimal execution path before starting work: plan-vs-act gates, agent architecture selection (chaining, routing, parallelization, orchestrator-worker, evaluator-optimizer), scope management heuristics, critical-path analysis across task networks, and context budget awareness. Use when deciding how to approach a task (plan first vs act immediately), decomposing complex work into parallelizable subtasks, choosing between subagent patterns, or when a task has failed twice and needs a fresh approach. Do NOT use for executing the chosen plan (use task-execution), debugging failures (use troubleshooting or diagnosis), or tool-level efficiency (use tool-call-strategy). Do NOT use for reduce the number of tool calls in this read-heavy step. Do NOT use for troubleshoot why the deployment keeps timing out. Do NOT use for execute the implementation plan we already agreed on.

jacob-balslev Skill

Use when reasoning about code coverage as a strategic measurement rather than a quality target: structural reach versus behavioral verification (covered vs tested), the coverage-criterion hierarchy (function, line, branch, decision, condition, MC/DC, path) and matching the criterion to where a module's failure modes hide, Marick's floor-vs-ceiling distinction, Goodhart's Law applied to coverage and Goodhart-resistant policy (diagnostic use, diff/patch floors, risk-weighting, behavioral panels), denominator hygiene and cross-tool counter semantics, safety-critical MC/DC (DO-178C Level A required; ISO 26262 ASIL-D highly recommended) as requirements-based evidence, and interpreting coverage from AI-generated tests. Do NOT use for choosing test levels (use testing-strategy), mutation testing as a behavioral signal (use mutation-testing), test-double construction (use test-doubles-design), LLM eval iteration (use eval-driven-development), or specific coverage-tooling configuration (tool docs).

jacob-balslev Skill

Use when facing a new problem and unsure which problem-solving methodology or foundational skill to apply first. Routes between first-principles-thinking, pattern-recognition, mental-models, constraint-awareness, and task-analysis by classifying the problem type. Activate before choosing any other foundational skill. Do NOT use to execute the selected approach (use the specific routed skill directly), for implementation work (use the relevant engineering skill), or when the correct approach is already known. Do NOT use for actually apply first-principles thinking to this specific problem. Do NOT use for analyze why this pattern keeps recurring in the codebase. Do NOT use for break down the constraints on this engineering decision. Do NOT use for map the mental model for this domain concept. Do NOT use for implement the algorithm we already decided on. Do NOT use for choose a testing strategy for this feature.

jacob-balslev Skill

Use when starting a non-trivial feature, refactor, or agent implementation that needs a written spec, plan, task breakdown, and verification path before code changes. Covers Spec Kit-style SDD phases, requirements-vs-plan separation, task traceability, review gates, and spec-compliance verification. Do NOT use for one-line edits, README-only fixes, post-implementation code review (use `code-review`), or test-level decisions (use `testing-strategy`).

jacob-balslev Skill

Use when reasoning about Test-Driven Development as a design discipline rather than a workflow: the red-green-refactor cycle as a feedback loop, the difference between London-school (outside-in, interaction-heavy, mock-driven) and Detroit-school (inside-out, state-heavy, classicist) TDD, the role of TDD as a design tool (how tests pressure code into more decomposable shapes), the connection between TDD and emergent design, the boundary between TDD and prior-test-suites, why TDD's failure mode is not 'no tests' but 'tests that mirror implementation', and the empirical record of TDD's effects on defect density, design quality, and development velocity. Do NOT use for the strategy of what to test at which level (use testing-strategy), the construction of test doubles (use test-doubles-design), the discipline of LLM eval iteration (use eval-driven-development), or general-software process workflow (use the obra/superpowers test-driven-development workflow skill — this skill is the concept-shape complement).

jacob-balslev Skill

Use when designing, reviewing, or debugging an autonomous AI agent loop: repeated agent execution, completion signals, checkpoints, supervisor respawn, stall detection, safety caps, and human handoff rules. Covers the core loop patterns from simple bounded runs through sentinel-based continuation, checkpoint-resume, and external supervisor loops. Do NOT use for choosing a specific agent product command (use agent-engineering or the product's docs), writing ordinary task instructions (use prompt-craft), or optimizing individual tool calls (use tool-call-strategy).

jacob-balslev Skill

Use when reasoning about systems that pass untrusted content to a language model: the data-vs-instruction collapse that makes this attack class a structural property of LLMs rather than a fixable bug, the direct/indirect/exfiltration/action-trigger taxonomy, the role of every untrusted surface (RAG retrievals, tool results, attachments, web content, document parsing, user-provided text), why content filters and improved system prompts do not solve it, and the defense-in-depth measures that do (capability constraint, content origin tracking, separate planning and execution stages, human-in-the-loop gates, principle-of-least-authority for tools). Do NOT use for model refusal policy or jailbreak evals (use `guardrails` or `eval-driven-development`), for general application security (use `owasp-security` or `security-fundamentals`), for runtime input validation patterns (use `type-safety` + `api-design`), or for the protocol cycle of tool calls (use `tool-call-flow`).

coderabbitai Skill

Vercel AI Gateway expert guidance. Use when configuring model routing, provider failover, cost tracking, or managing multiple AI providers through a unified API.

coderabbitai Skill

Vercel app and REST API expert guidance. Use when the agent needs live access to Vercel projects, deployments, environment variables, domains, logs, or documentation through the connected Vercel app or REST API.

jscraik Bundle

Assists with questions about Guy Podjarny's talk "Skills are the new Code". Use when the user wants to understand, apply, audit, or explore frameworks from this keynote — including the five engineering disciplines for skills (static analysis, evals, security testing, dependency management, observability), the three challenge buckets, the agentic development stack, or concepts like skill authoring, context engineering, agent harnesses, and skill quality scoring.

jscraik Bundle

Explains, summarizes, and turns Matthias Luebken's talk on embedding Pi-style coding agents into safe product-design artifacts: tool-contract sketches, guardrail checklists, session-record models, and malleable-software review plans. Use when the user asks about OpenClaw, Pi-style product agents, lifecycle guardrails, agent sessions, or design-level application of these primitives.

jscraik Bundle

Answers questions about Lamis's (Anthropic) AI Native DevCon talk on context engineering, agent memory systems, and dreaming — an asynchronous, out-of-band memory-curation process. Supports factual Q&A, framework application, system auditing, artifact drafting, and concept explanation based on the talk's content. Use when the user asks about context engineering, CLAUDE.md files, agent memory persistence, skills, multi-session memory, the dreaming process, hashing-based concurrency, or wants to apply, audit against, or draft artifacts from the frameworks described in this talk.

jscraik Bundle

Answers questions about, summarises key insights from, and helps apply concepts from Simon Martinelli's talk "Lessons from Spec-driven Development" — providing verbatim-grounded explanations, audits, and artifact drafts. Use when the user asks about the AI Unified Process, system use cases as specs (vs user stories), self-contained systems vs microservices, skills/MCP servers/guardrails, AI-assisted ERP modernization, drift management, how architecture style impacts AI coding agents, or applying his spec-driven approach to current work.

jscraik Bundle

Answers questions about, summarizes, and applies May Walter's AI Native DevCon talk "From Blind Spots to Merged PRs" on runtime intelligence for coding agents. Use when the user asks about production telemetry for agents, prod-to-code mapping, performance fixes from runtime data, why automated PRs need provenance, Hud's runtime code sensor, weekly performance reports, AI-generated fixes with production context, or applying Walter's evidence-first agent workflow to engineering teams.

stevenke1981 Skill

當使用者需要「成人內容合規創作者」處理成人娛樂業相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再在成年人、自願、合法與非剝削前提下處理內容或營運需求，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「成人情感文學作者」處理成人娛樂業相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再在成年人、自願、合法與非剝削前提下處理內容或營運需求，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「學術歷史學家」處理學術研究相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再建立來源透明、可反駁、可重現的研究分析，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「使用者體驗研究員」處理設計相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再把使用者需求轉成一致、可用、可存取且可實作的設計規格，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「站點可靠性工程師」處理工程研發相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再把需求轉成可實作、可測試、可回滾的工程方案，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「財務規劃與分析師」處理財務相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再建立可稽核的財務分析、情境假設、風險揭露與決策摘要，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「遊戲音訊工程師」處理遊戲開發相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再把玩法、內容、技術限制與玩家體驗轉成可測試的遊戲開發規格，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「商機策略師」處理銷售相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再建立以客戶需求、資格判定、價值證據與下一步為核心的銷售流程，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「階段 2：基礎建設 Agent」處理策略編排相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再建立範圍、責任、里程碑、風險、依賴與驗收清楚的執行計畫，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「社群內容創作者」處理媒體傳播相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再把目標、受眾與風格限制轉成可直接使用的創作交付物，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「網路犯罪研究員」處理爭議灰色行業相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再以辨識、預防、合規、教育與傷害降低為目的進行分析，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「宣傳戰研究分析師」處理極端組織分析相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再以辨識、預防、合規、教育與傷害降低為目的進行分析，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「社交工程防禦分析師」處理網路地下相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再以防禦、偵測、回應與風險降低為目的提供分析，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「氣候科技顧問」處理新興職業相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再建立來源透明、可反駁、可重現的研究分析，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「學術地理學家」處理學術研究相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再建立來源透明、可反駁、可重現的研究分析，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「品牌守護者」處理設計相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再把使用者需求轉成一致、可用、可存取且可實作的設計規格，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「Unreal 世界建構師」處理遊戲開發相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再把玩法、內容、技術限制與玩家體驗轉成可測試的遊戲開發規格，並輸出證據、風險、下一步與需要人工覆核的事項。

stevenke1981 Skill

當使用者需要「Auditor 專家」處理付費媒體相關任務時啟動。本 Agent 會先確認目標、資料來源、限制與驗收標準，再建立受眾、訊息、通路、實驗與衡量方法一致的成長方案，並輸出證據、風險、下一步與需要人工覆核的事項。

jscraik Bundle

Use when hardening, converting, auditing, or pre-release checking a Codex plugin package by verifying manifest paths, bundled skills, hooks, MCP/app config, validation gates, and release blockers.