ai-native-toolkit plugin
Skills for AI-native development: /assess (codebase readiness scoring + complexity hotspot SVG) and /huddle (Six Thinking Hats deliberation with Fibonacci team sizing). Also bundles user-invoked personal workflow skills (/tm, /issues, /fix-pr, /fix-develop).
How to install the ai-native-toolkit plugin
Run both commands inside Claude Code. The first adds the marketplace once; the second installs the plugin.
/plugin marketplace add bjcoombs/ai-native-toolkit
/plugin install ai-native-toolkit@ai-native-toolkitWhat's inside ai-native-toolkit
27 parts read from the plugin's own files.
Agents 8
Call as @agent-ai-native-toolkit:name- @agent-ai-native-toolkit:assess-layer-scorer
No description in the file.
- @agent-ai-native-toolkit:black-hat
No description in the file.
- @agent-ai-native-toolkit:blue-hat
No description in the file.
- @agent-ai-native-toolkit:green-hat
No description in the file.
- @agent-ai-native-toolkit:red-hat
No description in the file.
- @agent-ai-native-toolkit:scribe
No description in the file.
- @agent-ai-native-toolkit:white-hat
No description in the file.
- @agent-ai-native-toolkit:yellow-hat
No description in the file.
Skills 19
Claude loads it when the task matches- 6hats
Performs a Six Thinking Hats analysis by spawning specialized agents to evaluate a topic and synthesizing the results via a Blue Hat agent.
- ab-equivalence
Compares two versions of an LLM-directed document to determine if a candidate version maintains the same behavior as the original across a test set.
- assess
Assess codebase readiness for AI agents using a layered contract model and generate complexity treemaps and documentation navigability graphs.
- assess-findings
Assembles a codebase assessment report by combining deterministic data from run-context JSON and layer scorecards into a structured Markdown document.
- assess-pr
Automates post-assessment workflows by offering to open pull requests, track issues, freeze CI gates, and file tool feedback.
- deslop
Detects and removes AI-generated 'slop' by identifying patterns like puffery, superficial analysis, and repetitive rhetorical structures in written text.
- fix-develop
Automates the diagnosis and fixing of failing CI workflows on a repository's default branch by creating worktrees and pull requests.
- fix-pr
Iterates autonomously on pull request CI failures and review comments until all merge criteria are met.
- ghreport
Generates read-only health reports for GitHub and GitLab organizations, summarizing open PRs, CI status, security alerts, and branch protection.
- ghsync
Bulk clone and synchronize all accessible repositories from GitHub or GitLab into a structured worktree-friendly local layout.
- huddle
Performs structured multi-perspective analysis using the Six Thinking Hats methodology with professional lens agents.
- issues
Triages open GitHub issues by researching scope, checking for overlaps with PRs, and decomposing complex tasks into agent-ready sub-issues.
- marathon
Orchestrates autonomous multi-unit agent teams to complete work lists by managing dependency DAGs, hot-file mapping, and smart-merge waves.
- pr-review-merge
Automates the pull request lifecycle by verifying six merge criteria and performing a smart merge once all conditions are met.
- semantic-compress
Reduces the token count of LLM-directed documents by replacing core knowledge with pointers while preserving bespoke project details.
- skill-forge
Hardens agent instruction files and skills through adversarial judge-panel refinement rounds and a three-tier promotion gate.
- tm
An orchestrator that plans, executes, reviews, and closes development tasks using subagents, worktrees, and automated PR workflows.
- tm-marathon-config-example
Provides a configuration template for CLAUDE.md to manage marathon mode settings, bot reviewers, CI patterns, and GitHub issue labels.
- understand
Performs exhaustive context-gathering and problem definition to ensure deep understanding before proposing solutions.
Using ai-native-toolkit in Claude Code
Ask for an agent by name:
@agent-ai-native-toolkit:assess-layer-scorer
@agent-ai-native-toolkit:black-hatSkills need no command. Claude reads them when your request matches what they describe.
Update, turn off or remove
- Update the plugin (shell)
- claude plugin update ai-native-toolkit@ai-native-toolkit
- Refresh the marketplace
- /plugin marketplace update ai-native-toolkit
- Turn it off for now
- /plugin disable ai-native-toolkit@ai-native-toolkit
- Remove it
- /plugin uninstall ai-native-toolkit@ai-native-toolkit
Skills in ai-native-toolkit
- ab-equivalencebjcoombs/ai-native-toolkit
Compares two versions of an LLM-directed document to determine if a candidate version maintains the same behavior as the original across a test set.
4 files32stars - assessbjcoombs/ai-native-toolkit
Assess codebase readiness for AI agents using a layered contract model and generate complexity treemaps and documentation navigability graphs.
94 files32stars - deslopbjcoombs/ai-native-toolkit
Detects and removes AI-generated 'slop' by identifying patterns like puffery, superficial analysis, and repetitive rhetorical structures in written text.
2 files32stars - ghreportbjcoombs/ai-native-toolkit
Generates read-only health reports for GitHub and GitLab organizations, summarizing open PRs, CI status, security alerts, and branch protection.
2 files32stars - ghsyncbjcoombs/ai-native-toolkit
Bulk clone and synchronize all accessible repositories from GitHub or GitLab into a structured worktree-friendly local layout.
2 files32stars - marathonbjcoombs/ai-native-toolkit
Orchestrates autonomous multi-unit agent teams to complete work lists by managing dependency DAGs, hot-file mapping, and smart-merge waves.
3 files32stars - semantic-compressbjcoombs/ai-native-toolkit
Reduces the token count of LLM-directed documents by replacing core knowledge with pointers while preserving bespoke project details.
8 files32stars - skill-forgebjcoombs/ai-native-toolkit
Hardens agent instruction files and skills through adversarial judge-panel refinement rounds and a three-tier promotion gate.
6 files32stars - assess-findingsbjcoombs/ai-native-toolkit
Assembles a codebase assessment report by combining deterministic data from run-context JSON and layer scorecards into a structured Markdown document.
32stars - assess-prbjcoombs/ai-native-toolkit
Automates post-assessment workflows by offering to open pull requests, track issues, freeze CI gates, and file tool feedback.
32stars - fix-developbjcoombs/ai-native-toolkit
Automates the diagnosis and fixing of failing CI workflows on a repository's default branch by creating worktrees and pull requests.
32stars - fix-prbjcoombs/ai-native-toolkit
Iterates autonomously on pull request CI failures and review comments until all merge criteria are met.
32stars - huddlebjcoombs/ai-native-toolkit
Performs structured multi-perspective analysis using the Six Thinking Hats methodology with professional lens agents.
32stars - issuesbjcoombs/ai-native-toolkit
Triages open GitHub issues by researching scope, checking for overlaps with PRs, and decomposing complex tasks into agent-ready sub-issues.
32stars - pr-review-mergebjcoombs/ai-native-toolkit
Automates the pull request lifecycle by verifying six merge criteria and performing a smart merge once all conditions are met.
32stars - tm-marathon-config-examplebjcoombs/ai-native-toolkit
Provides a configuration template for CLAUDE.md to manage marathon mode settings, bot reviewers, CI patterns, and GitHub issue labels.
32stars - understandbjcoombs/ai-native-toolkit
Performs exhaustive context-gathering and problem definition to ensure deep understanding before proposing solutions.
32stars - tmbjcoombs/ai-native-toolkit
An orchestrator that plans, executes, reviews, and closes development tasks using subagents, worktrees, and automated PR workflows.
32stars - 6hatsbjcoombs/ai-native-toolkit
Performs a Six Thinking Hats analysis by spawning specialized agents to evaluate a topic and synthesizing the results via a Blue Hat agent.
32stars
ai-native-toolkit plugin questions
How do I install the ai-native-toolkit plugin?
In Claude Code run /plugin marketplace add bjcoombs/ai-native-toolkit, then /plugin install ai-native-toolkit@ai-native-toolkit.
What does the ai-native-toolkit plugin include?
ai-native-toolkit bundles 8 agents, 19 skills.
Which agents come with ai-native-toolkit?
assess-layer-scorer, black-hat, blue-hat, green-hat, red-hat, scribe +2. Call one with @agent-ai-native-toolkit:assess-layer-scorer.
How do I update or remove the ai-native-toolkit plugin?
Update it with claude plugin update ai-native-toolkit@ai-native-toolkit in your shell, or from the Installed tab of /plugin. Remove it with /plugin uninstall ai-native-toolkit@ai-native-toolkit.