agent-toolkit-cli 1.0.3__tar.gz

agent_toolkit_cli-1.0.3/.gitignore

@@ -0,0 +1,25 @@
+# Python
+__pycache__/
+*.py[cod]
+.venv/
+*.egg-info/
+
+# Node
+node_modules/
+
+# OS
+.DS_Store
+*.swp
+
+# Secrets - never commit
+*.env
+.env.*
+!.env.example
+
+# Build artifacts
+dist/
+build/
+
+# Editor
+.idea/
+.vscode/settings.json

agent_toolkit_cli-1.0.3/AGENTS.md

@@ -0,0 +1,304 @@
+# AGENTS.md — AI Agent Contract
+
+**Purpose**: This file is the primary contract for any AI agent contributing to or working inside this repository. Read it before touching any file.
+
+---
+
+## What This Toolkit Does
+
+`agent-toolkit` is a modular library of AI capabilities — skills, agent personas, loop templates, MCP configs, and per-tool profiles — designed to work across every major AI coding assistant (Claude Code, Cursor, OpenCode, GitHub Copilot, Windsurf, Pi Coding Agent).
+
+The goal is a single source of truth: author a skill once, deploy it to any tool via its profile. When you add, modify, or remove anything here, you are affecting every downstream consumer who installs this toolkit.
+
+**Operate accordingly: be precise, cite sources, and validate before finalizing.**
+
+---
+
+## Repository Structure
+
+```
+agent-toolkit/
+├── skills/
+│   ├── core/           # Memory, planning, context injection
+│   ├── delivery/       # Code review, PR workflows, CI fix
+│   ├── design/         # UI/UX, design systems, components
+│   ├── forge/          # Code generation, TDD, refactoring
+│   ├── integrations/   # JIRA, Confluence, Slack, Linear, ClickUp, Notion
+│   ├── data/           # DBT, Snowflake, data pipelines
+│   ├── tooling/        # Git, Docker, CI/CD, env setup
+│   ├── ops/            # Incident, security review, performance
+│   └── loops/          # Recurring agentic workflow skills
+├── agents/             # Tool-agnostic agent persona definitions
+├── profiles/
+│   ├── claude-code/    # .claude/settings.json overlay + CLAUDE.md
+│   ├── cursor/         # .mdc rule files for Cursor
+│   ├── opencode/       # System prompt overlays for OpenCode
+│   ├── copilot/        # copilot-instructions.md for GitHub Copilot
+│   ├── windsurf/       # rules.md + memory files for Windsurf
+│   └── pi/             # Skill files for Pi Coding Agent
+├── mcp/
+│   └── templates/      # MCP server config templates (JSON)
+├── loops/              # Loop engineering templates
+│   └── <name>/
+│       ├── request.md  # Prompt template with YAML frontmatter
+│       ├── report.md   # Output report template
+│       └── runbook.md  # Human operational runbook
+├── packs/              # Solution packs (YAML bundles)
+├── catalogs/
+│   ├── skill-catalog.yaml
+│   └── agent-catalog.yaml
+├── schemas/            # JSON schemas for validation
+│   ├── skill.schema.json
+│   └── loop.schema.json
+├── docs/               # Human-facing documentation
+└── scripts/            # install.sh, validate-skills.sh, validate-loops.sh
+```
+
+---
+
+## Operating Rules
+
+**Always:**
+
+- Check `catalogs/skill-catalog.yaml` before adding a new skill — avoid duplicates
+- Validate with `bash scripts/validate-skills.sh` before marking any skill work complete
+- Cite which file a convention comes from when referencing repo standards
+- Use English for all file content, commit messages, and PR descriptions
+- Check the compatibility matrix in `skill.json` before claiming a skill works with a tool
+- Keep secrets out of the repository — MCP templates use placeholder values only
+
+**Never:**
+
+- Commit credentials, API keys, or tokens to any file
+- Claim a skill is compatible with a tool without testing or explicit authorship confirmation
+- Skip the `skill.json` manifest when adding a skill
+- Modify `catalogs/` manually — these are generated by `scripts/build-catalog.sh`
+- Add content specific to any private organization — this is a public, vendor-neutral toolkit
+
+---
+
+## How to Add a Skill
+
+Each skill lives in its own directory under the appropriate domain:
+
+```
+skills/<domain>/<skill-name>/
+├── SKILL.md       # Human-readable description and usage guide
+└── skill.json     # Machine-readable manifest
+```
+
+### SKILL.md Frontmatter Spec
+
+Every `SKILL.md` must begin with a YAML frontmatter block:
+
+```markdown
+---
+name: my-skill-name
+description: One-sentence description of what this skill does.
+metadata:
+  author: github-username
+  version: "1.0.0"
+  tags:
+    - tag1
+    - tag2
+  domain: delivery        # one of: core, delivery, design, forge, integrations, data, tooling, ops, loops
+---
+
+# My Skill Name
+
+[Human-readable usage guide follows]
+```
+
+Required frontmatter fields:
+
+| Field | Type | Description |
+|---|---|---|
+| `name` | string | Kebab-case skill identifier, unique within the domain |
+| `description` | string | One sentence, present tense ("Reviews pull requests for…") |
+| `metadata.author` | string | GitHub username of the primary author |
+| `metadata.version` | string | Semver string, start at `"1.0.0"` |
+| `metadata.tags` | list | Searchable tags, lowercase kebab-case |
+| `metadata.domain` | string | Parent domain directory name |
+
+### skill.json Spec
+
+Every skill directory must also contain a `skill.json` manifest:
+
+```json
+{
+  "name": "my-skill-name",
+  "version": "1.0.0",
+  "description": "One-sentence description matching SKILL.md.",
+  "source": "skills/delivery/my-skill-name",
+  "author": "github-username",
+  "tags": ["tag1", "tag2"],
+  "compatibility": {
+    "claude-code": true,
+    "cursor": true,
+    "opencode": true,
+    "copilot": false,
+    "windsurf": true,
+    "pi": false
+  }
+}
+```
+
+`compatibility` values:
+
+- `true` — skill works with this tool and has been verified
+- `false` — skill does not apply to this tool (e.g., tool-specific API unavailable)
+- `"partial"` — skill works with limitations; document them in `SKILL.md`
+
+The `skill.json` is validated against `schemas/skill.schema.json`. Run `bash scripts/validate-skills.sh` to check.
+
+---
+
+## How to Add a Loop Template
+
+Loop templates live in `loops/<loop-name>/`. A complete loop template requires three files.
+
+### request.md — Prompt Template
+
+```markdown
+---
+name: my-loop-name
+tier: L1                        # L1 (minutes-hours), L2 (daily), L3 (weekly/monthly)
+cadence: "*/30 * * * *"         # cron expression or human description
+goal: "One sentence describing the loop's objective."
+allowlist:
+  - github
+  - slack
+deny:
+  - file_write                  # tools the loop must never invoke
+budget:
+  tokens: 50000                 # max tokens per run
+  duration_minutes: 10          # hard timeout
+resumable: true                 # whether partial runs can be resumed
+---
+
+[Prompt body follows. Use {{variables}} for substitution at runtime.]
+```
+
+Required frontmatter fields:
+
+| Field | Type | Description |
+|---|---|---|
+| `name` | string | Kebab-case loop identifier |
+| `tier` | string | `L1`, `L2`, or `L3` |
+| `cadence` | string | Cron expression or human-readable cadence |
+| `goal` | string | One sentence describing what the loop achieves |
+| `allowlist` | list | MCP servers or tool categories the loop may use |
+| `deny` | list | Capabilities explicitly forbidden in this loop |
+| `budget.tokens` | integer | Maximum token budget per run |
+| `budget.duration_minutes` | integer | Hard timeout in minutes |
+| `resumable` | boolean | Whether the loop supports mid-run checkpoint/resume |
+
+### report.md — Output Template
+
+Define the structure of the loop's output report. Use Markdown with placeholder sections that the loop fills in at runtime.
+
+### runbook.md — Operational Runbook
+
+Human-readable guide covering:
+- What the loop does and why
+- How to trigger it manually
+- How to interpret its output
+- Common failure modes and remediation steps
+- Escalation path if the loop cannot complete
+
+---
+
+## How to Add a Profile
+
+Profiles adapt toolkit skills for a specific tool. Each profile lives in `profiles/<tool-name>/`.
+
+### Claude Code (`profiles/claude-code/`)
+
+- `CLAUDE.md` — system prompt overlay referencing skills by path
+- `settings.json` — skill loading configuration
+
+### Cursor (`profiles/cursor/`)
+
+- `rules/<domain>.mdc` — one rule file per domain, using Cursor's MDC format
+- Keep each rule file focused on one domain; avoid catch-all files
+
+### OpenCode (`profiles/opencode/`)
+
+- `system-prompt.md` — system prompt that opencode loads on startup
+- `skills.yaml` — list of skill paths to inject
+
+### GitHub Copilot (`profiles/copilot/`)
+
+- `copilot-instructions.md` — the file that gets copied to `.github/copilot-instructions.md` in projects
+- Must be self-contained (Copilot cannot reference external files at runtime)
+
+### Windsurf (`profiles/windsurf/`)
+
+- `rules.md` — rules file that Windsurf loads from `~/.codeium/windsurf/`
+- `memories/` — optional pre-seeded memory files
+
+### Pi Coding Agent (`profiles/pi/`)
+
+- `skills/<skill-name>.md` — individual skill files in Pi's expected format
+- Pi skills are standalone; each file must be self-contained
+
+---
+
+## Agent Persona Format
+
+Agent personas in `agents/` follow this structure:
+
+```
+agents/<persona-name>/
+├── README.md      # Role description, triggers, handoff rules
+└── persona.json   # Machine-readable manifest
+```
+
+`persona.json` fields:
+
+```json
+{
+  "name": "persona-name",
+  "role": "Short role title",
+  "description": "What this persona does and when to use it.",
+  "triggers": ["keyword1", "keyword2"],
+  "handoff_to": ["other-persona"],
+  "skills": ["skill-domain/skill-name"],
+  "compatibility": {
+    "claude-code": true,
+    "cursor": true,
+    "opencode": true,
+    "copilot": false,
+    "windsurf": true,
+    "pi": false
+  }
+}
+```
+
+---
+
+## Catalog Generation
+
+`catalogs/skill-catalog.yaml` and `catalogs/agent-catalog.yaml` are **generated files**. Do not edit them by hand.
+
+To regenerate after adding or modifying skills or agents:
+
+```bash
+bash scripts/build-catalog.sh
+```
+
+The CI pipeline runs this automatically and will fail if the catalogs are out of sync with the source files.
+
+---
+
+## Validation
+
+Before marking any contribution complete:
+
+```bash
+bash scripts/validate-skills.sh    # Validates all skill.json against schema
+bash scripts/validate-loops.sh     # Validates all loop frontmatter against schema
+bash scripts/build-catalog.sh      # Regenerates and checks catalogs
+```
+
+All three must exit 0.

agent_toolkit_cli-1.0.3/CHANGELOG.md

@@ -0,0 +1,63 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+---
+
+## [1.0.1] — 2026-08-04
+
+### Fixed
+- Publish agent-toolkit-cli to PyPI (release.yml was missing build + publish steps in v1.0.0)
+- SVG animations: replace CSS @keyframes with native SVG <animate> (GitHub strips CSS)
+- Banner SVG: fix text overlap between title and right panel (shift from x=575 to x=530)
+- CLI wheel install: shared _paths.py resolves toolkit root correctly in all install modes
+- CI: remove pytest || true so test failures block the pipeline
+
+## [1.0.2] — 2026-08-04
+
+### Fixed
+- release.yml now correctly builds wheel and publishes to PyPI
+
+## [1.0.3] — 2026-08-04
+
+### Fixed
+- release.yml YAML syntax fixed for GitHub Actions validator
+
+## [Unreleased]
+
+### Added
+
+- Initial toolkit release with 53+ skills across 9 domains (core, delivery, design, forge, integrations, data, tooling, ops, loops)
+- Agent personas for all major AI coding tools: architect, planner, code-reviewer, security-reviewer, performance-optimizer, tdd-guide, refactor-cleaner, docs-lookup
+- Profiles for Claude Code, Cursor, OpenCode, GitHub Copilot, Windsurf, and Pi Coding Agent
+- 10 loop engineering templates including OSS maintenance patterns:
+  - `oss-pr-monitor` (L1) — monitor open PRs across OSS repos
+  - `oss-triage` (L1) — triage new issues and apply labels
+  - `oss-daily-briefing` (L2) — daily activity summary across tracked repos
+  - `dependency-drift` (L2) — detect and PR outdated dependencies
+  - `ci-health` (L1) — watch CI status and auto-diagnose failures
+  - `release-notes` (L3) — draft release notes from merged PRs
+  - `security-sweep` (L2) — vulnerability scan across repos
+  - `codeowner-review` (L2) — remind code owners of pending reviews
+  - `stale-branch-cleanup` (L3) — identify and archive stale branches
+  - `contributor-digest` (L3) — generate weekly contributor activity digest
+- MCP configuration templates for 6 services: GitHub, Slack, Notion, Linear, Figma, ClickUp
+- Solution packs: `oss-ecosystem`, `startup-delivery`, `enterprise-ops`, `data-platform`
+- JSON schemas for skill and loop validation (`schemas/skill.schema.json`, `schemas/loop.schema.json`)
+- `catalogs/skill-catalog.yaml` — machine-readable index of all skills
+- `catalogs/agent-catalog.yaml` — machine-readable index of all agent personas
+- `scripts/validate-skills.sh` — validates all skill.json manifests against schema
+- `scripts/validate-loops.sh` — validates all loop frontmatter against schema
+- `scripts/build-catalog.sh` — regenerates skill and agent catalogs
+- `scripts/install.sh` — quick-setup installer for all supported tools
+- `AGENTS.md` — AI agent contract with authoring specs for skills, loops, and profiles
+- `CONTRIBUTING.md` — contributor guide with step-by-step instructions
+- `SECURITY.md` — security policy including private disclosure process
+- GitHub Actions CI workflow for automated skill and loop validation on every push and PR
+
+---
+
+[Unreleased]: https://github.com/ulises-jeremias/agent-toolkit/compare/HEAD...HEAD

agent_toolkit_cli-1.0.3/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Ulises Jeremias
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.