Add instructions for enabling skills via YAML frontmatter. Clarify that
skills are disabled by default to prevent context pollution and specify
default skill sets for the Plugin Creator, Configurator, and Security
Reviewer agents.
Apply research-backed formatting across all agents, commands, and
plugins to improve LLM precision. This update also replaces the
commands-md-updater agent with a new /sync-commands workflow.
- Add Security Reviewer with 10 new security skills.
- Split Vite-React expert into orchestrator and brain roles.
- Add /init and /refactor-rfc-xml commands.
- Update model references to GPT-5.2 and remove deprecated agents.
Update model researcher and schema references to target Q4 2025 frontier
models. Add a refactor-rfc-xml command to restructure markdown files and
reorganize the AGENTS.md directory structure.
Apply a default deny policy for skills across all agents. Refine
frontmatter descriptions for clarity and restructure the
commands-md-updater agent using semantic tags. Update the TypeScript
expert to include test writing capabilities.
Remove CLAUDE.md and mark Commands2Skills as partially deprecated to
reflect OpenCode's native Skills support. Add VRTCT-orchestrator agent
and an init command for AGENTS.md. Improve template documentation.
Convert markdown files to use XML-like tags for structure and adopt RFC
2119 keywords. Delete redundant generic agent definitions and update
coding guidelines to enforce stricter modularity, file size limits, and
barrel exports.
Introduce a standardized style using RFC 2119 normative keywords and XML
tags to improve LLM prompt structure and parsing.
Includes a new style guide, an automated refactoring command, and
updates to existing agent and skill documentation.
Adds skill to find, install, and configure MCP servers Includes catalog
of 16 common MCP servers Provides configuration examples and
troubleshooting guide
Configure /improve:save to store prompts in an `enhanced-prompts`
directory and run as a subtask. Document the new /refactor command
focused on modularity and cleanup.
Add CODING-TS.MD with coding architecture principles. Require plugins to
use subdirectories (plugin-name/index.ts) and split complex plugins into
multiple files under 150 lines each.
Adds a "Before Publishing" section to prompt the user for package name,
scope, version, and license. Updates package.json and README examples to
use placeholders instead of hardcoded values. Includes instructions for
scoped package publishing.
Clarify that OpenCode auto-installs dependencies and remove npm install
instructions. Update package.json to use peerDependencies and include
example config files. Restrict agent permissions.
- Introduces a new build agent with "no fluff" enhancement approach
- Appends strict implementation standards and intentional minimalism
- Updated README.md to include the new agent in the collection
- Added build.md to the Generic agent's configuration directory
Add mandatory reading requirements to all agents and introduce
CODING-TS.md for universal coding standards. Update the main agent to
prioritize delegation to specialized subagents.
- Add /improve:run command for immediate prompt enhancement and execution
- Add /improve:save command for prompt enhancement with review workflow
- Both commands integrate official prompt engineering guides from:
* OpenAI GPT-5.1 & GPT-5.1 Codex
* Anthropic Claude 4.5 Sonnet & Opus
* Google Gemini 3 Pro
- Update README.md with new command descriptions
Test User