claude

devkit-ai-engineering

Name: devkit-ai-engineering
Author: kaue rodrigues dos santos

Building with LLMs: evaluation, multi-agent patterns, agent tool and CLI design, context compression and skill vetting.

0 on GitHubby kaue rodrigues dos santosv0.1.1MITupdated yesterdaydevelopment

Plugin devkit-ai-engineering from marketplace devkit in chacal88/devkit. Commands read from the repo's manifest — today.

What it installs

Skills (6)

agent-friendly-cli
Design spec (98 rules, three certification levels) for CLI tools that AI agents can use safely - structured JSON output, error handling, input contracts, safety guardrails, exit codes and self-description. Use when building or modifying a CLI that agents will call, or when the user asks to make a CL
agent-tool-design
Designs tools that agents use effectively - clear descriptions and schemas, consolidation and architectural reduction, error messages agents can act on. Use when creating tools or MCP servers for an agent system, debugging tool misuse or tool-related failures, or trimming an existing tool set.
context-compression
Strategies for compressing long agent context - anchored iterative summarization, opaque compression, regenerative summaries - optimized for tokens per task rather than tokens per request, plus probe-based evaluation of compression quality. Use when agent sessions exceed the context window, designin
llm-evaluation
Builds LLM evaluation - LLM-as-judge with direct scoring and pairwise comparison, rubrics, and mitigation of position, length and authority bias in evaluation pipelines. Use when the user asks to implement LLM-as-judge, compare model outputs, create evaluation rubrics, mitigate evaluation bias, or a
multi-agent-patterns
Patterns for multi-agent systems - supervisor/orchestrator, peer-to-peer swarm, hierarchical - with context isolation, handoffs, sub-agents and parallel execution, and their failure modes. Use when the user asks to design a multi-agent system, implement a supervisor or swarm, coordinate multiple age
skill-scanner
Scans an agent skill for security issues before adoption - prompt injection, malicious code, excessive permissions, secret exposure and supply-chain risk - with a manual review checklist. Use when evaluating a third-party skill or plugin before installing it, or when the user asks whether a skill is

View source on GitHub

More from chacal88/devkit

All 6 plugins →

devkit-core

claude

Stack-agnostic dev workflow: research, plan, implement, review, ship, debug and test.

8 skills · 14 agents

devkit-stack

claude

Technology references: Postgres, Neon, React performance, Vitest, multi-tenant SaaS and threat modeling.

6 skills

devkit-design

claude

Greenfield design in four DDD phases: business discovery, domain model, architecture and implementation plan.

5 skills

Runs the devkit flow on top of OpenSpec: the devkit and devkit-lite schemas (grill → plan → readiness → TDD → review → evidence) and hooks that keep product code closed until readiness is APPROVED and refuse a PR without an approved review and evidence.

3 skills · 1 agent · hooks

devkit-panel

claude

Side pane in the Claude desktop app over the devkit checkers: the change this session conducts with its next step, phases, PR, delivery items and child sessions, the other changes in progress, and buttons that put the owner's next command in the prompt box. Read-only.

Similar plugins

superpowers

claudecodex

Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques

15 skills · hooks

obra/superpowers 316,836

career-ops

claudecodex

Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications

5 skills · hooks

career-ops-hq/career-ops 82,253

context7

claudecodex

Up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your coding agent's context.

1 MCP server

upstash/context7 64,709

cli-anything

claude

Build powerful, stateful CLI interfaces for any GUI application using the cli-anything harness methodology.

5 commands

HKUDS/CLI-Anything 58,665

academic-research-skills

claude

5 skills + 35 modes + Material Passport pipeline. Includes v3.6.7 cross-model audit gate and v3.6.8 generator-evaluator contract.

5 skills · 16 commands · 4 agents · hooks

Imbad0202/academic-research-skills 56,698

agent-browser

claude

Automates browser interactions for web testing, form filling, screenshots, and data extraction

12 skills

vercel-labs/agent-browser 50,373

Maintain devkit-ai-engineering?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex plugins6 plugins