agency-agents-ja
404体の即戦力AI専門エージェント集 — Claude Code / Cursor / Copilot 等に対応。日本市場向けオリジナル125体(SIer・受託開発・SaaS・製造DX・公共分野)。agency-agentsの日本語コミュニティ版
- 10
- Shell
- MIT
- 9 h ago
sscodeai
Platform Engineer in Japan · Model-agnostic AI delivery platforms · agency-agents-ja maintainer
Built from sscodeai's public GitHub profile, Japan. Not on Take My Repo yet, so there are no posts or updates here.
404体の即戦力AI専門エージェント集 — Claude Code / Cursor / Copilot 等に対応。日本市場向けオリジナル125体(SIer・受託開発・SaaS・製造DX・公共分野)。agency-agentsの日本語コミュニティ版
Model-agnostic platform for verifiable agentic software delivery: workflows, traceability, evidence-native testing, agent evals, and verify-before-done loops.
AI Skill Maker is a meta-skill for creating and refreshing reusable AI skills and assistant instruction bundles across platforms.
日本のIT開発現場向けSuperpowers。AI coding agentに計画、TDD、レビュー、検証、Backlog/Redmine/Jiraを前提にした日本語ワークフローを追加します。
The keysmith for AI agents: Go MCP server + CLI for agent-safe secrets with age encryption, masked views, handle passing, self-healing rotation, and Vault short-TTL credentials.
Demonstration target for takumi's black-box delivery mode: a confirmed requirement becomes a delivered product through an independently forged judge, a pull request, an issue, a pipeline, and a merge.
AI Input Trust Gateway — zero-trust input layer: audit documents, source files and agent rule files for hidden prompt-injection before they reach an LLM/Agent
Delivery target for the takumi unattended chain — a FRESH repository, so a delivery must be implemented from scratch rather than verified from history
Explain like I am 5: picture-first single-file HTML explainers, any model, multi-language. /eli5 skill + template + export + examples.
Use a heterogeneous multi-model committee to expand the option space before deciding — human makes the final call. Agent Skills format (SKILL.md), works with Claude Code / Cursor / Codex / Gemini CLI.
Expand the decision space before you decide — multi-model committee + human final call. Eval: single-model 26% → naive multi-agent 78% → workbench 100% (12/12 hidden gold).
LLM API caching middleware that combines exact-match, semantic, and upstream prefix-cache accounting for OpenAI-compatible providers.
Agent Eval tool: score, audit & explain agent trajectories — metrics + causal failure attribution in pure Go.
Generate project maintainer skills and instructions for AI coding agents.
Japanese localization of sscodeai/skills for native Japanese developers