Conversation
…tagent#20) Switch claude-code adapter from loadAllSkills() to loadAllSkillMetadata() so CLAUDE.md only inlines skill name, description, and allowed-tools with a file path reference instead of full SKILL.md instructions. Claude Code can read the full instructions on demand, saving ~5000 tokens per skill. Closes open-gitagent#19 Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Add deterministic multi-step workflow pattern matching the landing page, with YAML example showing skill/agent/tool step chaining. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Hullo 👋 @shreyas-lyzr I ran your skills through `tessl skill review` at work and found some targeted improvements. Here's the full before/after: | Skill | Before | After | Change | |-------|--------|-------|--------| | regulatory-analysis | 44% | 94% | +50% | | code-review | 53% | 94% | +41% | | export-agent | 55% | 94% | +39% | | research | 55% | 89% | +34% | | document-review | 49% | 79% | +30% | | create-agent | 60% | 90% | +30% | | run-agent | 60% | 90% | +30% | | get-started | 66% | 95% | +29% | | manage-skills | 66% | 94% | +28% | ## What Improved all 9 example skills across the `examples/` directory — descriptions, trigger terms, concrete examples, and workflow clarity. ## Why The skill descriptions lacked explicit "Use when..." clauses and specific trigger terms, making it harder for agents to select the right skill. Content was solid but missing concrete examples and validation steps in several cases. <details> <summary>Changes made</summary> **Description improvements (all 9 skills):** - Added explicit "Use when..." clauses with natural trigger keywords - Expanded descriptions with specific concrete actions and domain terms - Switched from vague "Help users..." to third-person action verbs - Added specific framework/tool names as trigger terms **Content improvements:** - `document-review`: Added structured output format with example finding (FINRA 2210 violation) - `regulatory-analysis`: Added citation validation step and example finding with rule reference - `create-agent`: Replaced "When to Use" with actionable Quick Start workflow (5 numbered steps) - `export-agent`: Replaced "When to Use" with export verification step - `run-agent`: Added troubleshooting section for common authentication errors - `research`: Added complete example output demonstrating expected quality - `code-review`: Added example finding with before/after code fix (SQL injection) - `manage-skills`: Added installation verification step - `get-started`: Improved description specificity (content already scored 100%) </details> ## How Tested - [x] `tessl skill review` passed on all 9 skills with score improvements - [x] YAML frontmatter validates correctly (all skills pass validation checks) - [x] No changes to agent.yaml files or directory structure ## Checklist - [x] My code follows the existing style of this project - [x] I have updated documentation (if applicable) - [x] I have read the [CONTRIBUTING.md](../CONTRIBUTING.md) --- Honest disclosure — I work at @tesslio where we build tooling around skills like these. Not a pitch - just saw room for improvement and wanted to contribute. Want to self-improve your skills? Just point your agent (Claude Code, Codex, etc.) at [this Tessl guide](https://docs.tessl.io/evaluate/optimize-a-skill-using-best-practices) and ask it to optimize your skill. Ping me - [@popey](https://github.com/popey) - if you hit any snags. Thanks in advance 🙏
Add `copilot` format to `gitagent export` that generates `.github/agents/<name>.agent.md` and `.github/skills/<name>/SKILL.md` files compatible with GitHub Copilot CLI. The adapter maps gitagent's SOUL.md, RULES.md, DUTIES.md, skills, knowledge, and compliance constraints into Copilot's agent.md format with YAML frontmatter. Closes open-gitagent#18 Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
) Closes open-gitagent#6. Adds docs/comparison.md comparing gitagent against: - ADL and similar emerging agent definition formats - Raw YAML/JSON config files - Framework-native (inline code) approaches such as LangChain, CrewAI, AutoGen Covers portability, versioning, compliance, skills ecosystem, and multi-agent topology dimensions. Includes migration guidance and hybrid patterns.
- Export adapter (src/adapters/opencode.ts): generates .opencode/instructions.md and opencode.json with provider/model config from agent.yaml - Runner (src/runners/opencode.ts): creates temp workspace with instructions + config, launches `opencode` CLI in interactive or single-shot mode - Import (src/commands/import.ts): reads .opencode/instructions.md + opencode.json, produces agent.yaml, SOUL.md, and RULES.md - Git runner auto-detection: detects .opencode/ or opencode.json and routes to the opencode adapter - Wired into export, run, and import commands with updated help text and README Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Use AGENTS.md instead of .opencode/instructions.md (per OpenCode docs) - Fix opencode.json config: model uses provider/model format, provider is object with npm package - Fix runner CLI: use `opencode run --prompt` for single-shot mode - Fix process.exit() before finally cleanup — move cleanup before exit - Extract shared buildComplianceSection to src/adapters/shared.ts (dedup from copilot.ts) - Fix importer to read AGENTS.md and parse provider/model format - Fix git auto-detection to check opencode.json instead of .opencode/ Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.