feat: add agent-skills ECC bundle#2
Open
ecc-tools[bot] wants to merge 11 commits into
Open
Conversation
…ited/agent-skills-instincts.yaml)
Reviewer's GuideAdds an auto-generated ECC bundle for the agent-skills repo, including Claude and Codex skills, Codex multi-agent configuration, and continuous-learning instincts derived from repository analysis. Sequence diagram for Codex reviewer agent using repo skill and MCPsequenceDiagram
actor Developer
participant CodexCLI
participant Config as Codex_config.toml
participant Reviewer as Reviewer_agent
participant RepoSkill as Repo_skill_agent_skills
participant GithubMCP as MCP_github
Developer->>CodexCLI: Invoke reviewer agent
CodexCLI->>Config: Load multi_agent configuration
Config-->>CodexCLI: Agent and MCP settings
CodexCLI->>Reviewer: Start session with sandbox_mode read-only
Reviewer->>RepoSkill: Apply repo patterns and workflows
RepoSkill-->>Reviewer: Conventions and commands
Reviewer->>GithubMCP: Fetch PR diff and repository context
GithubMCP-->>Reviewer: PR files, commits, metadata
Reviewer-->>Developer: Review focused on correctness, security, tests
File-Level Changes
Tips and commandsInteracting with Sourcery
Customizing Your ExperienceAccess your dashboard to:
Getting Help
|
![sourcery-ai[bot]](https://avatars.githubusercontent.com/in/48477?s=60&v=4){kind=small}
There was a problem hiding this comment.
Hey - I've found 2 issues, and left some high level feedback:
- Both SKILL.md files are wrapped in a top-level ```markdown code fence, which will prevent them from rendering as actual markdown in many viewers; consider removing the outer fence so the content is rendered normally.
- The instincts YAML includes overlapping rules for similar behaviors (e.g., file naming, import/export style, commit messages) with slightly different ids and triggers; it may be worth deduplicating or consolidating these to avoid conflicting guidance at runtime.
- In
agent-skills-instincts.yaml, thetriggerandsource_repofields are inconsistently formatted (e.g., capitalized vs lowercase triggers, URL vsorg/repostrings); aligning these formats will make the metadata more uniform and easier to process.
Prompt for AI Agents
Please address the comments from this code review:
## Overall Comments
- Both SKILL.md files are wrapped in a top-level ```markdown code fence, which will prevent them from rendering as actual markdown in many viewers; consider removing the outer fence so the content is rendered normally.
- The instincts YAML includes overlapping rules for similar behaviors (e.g., file naming, import/export style, commit messages) with slightly different ids and triggers; it may be worth deduplicating or consolidating these to avoid conflicting guidance at runtime.
- In `agent-skills-instincts.yaml`, the `trigger` and `source_repo` fields are inconsistently formatted (e.g., capitalized vs lowercase triggers, URL vs `org/repo` strings); aligning these formats will make the metadata more uniform and easier to process.
## Individual Comments
### Comment 1
<location path=".codex/config.toml" line_range="6" />
<code_context>
+# ECC Tools generated Codex baseline
+approval_policy = "on-request"
+sandbox_mode = "workspace-write"
+web_search = "live"
+
+[mcp_servers.github]
</code_context>
<issue_to_address>
**🚨 question (security):** Reconfirm that `web_search = "live"` is appropriate from a security/privacy standpoint for this project.
Because live web search increases the risk of data exfiltration and unpredictable external inputs, please confirm this repo will not handle sensitive or proprietary code, or add guardrails (e.g., limiting when/where web search is enabled) to align with your security posture.
</issue_to_address>
### Comment 2
<location path=".codex/config.toml" line_range="16" />
<code_context>
+command = "npx"
+args = ["-y", "@upstash/context7-mcp@latest"]
+
+[mcp_servers.exa]
+url = "https://mcp.exa.ai/mcp"
+
</code_context>
<issue_to_address>
**question (bug_risk):** Validate that the `mcp_servers.exa` configuration using only a `url` matches the expected MCP server schema.
`exa` is configured only with a `url`, whereas the others use `command`/`args`. If the MCP runtime expects the same shape for all entries or special handling for HTTP servers, this could cause runtime issues or the server being ignored. Please confirm this format against the current `config-schema.json` and the Exa MCP docs.
</issue_to_address>Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.
| # ECC Tools generated Codex baseline | ||
| approval_policy = "on-request" | ||
| sandbox_mode = "workspace-write" | ||
| web_search = "live" |
There was a problem hiding this comment.
🚨 question (security): Reconfirm that web_search = "live" is appropriate from a security/privacy standpoint for this project.
Because live web search increases the risk of data exfiltration and unpredictable external inputs, please confirm this repo will not handle sensitive or proprietary code, or add guardrails (e.g., limiting when/where web search is enabled) to align with your security posture.
| command = "npx" | ||
| args = ["-y", "@upstash/context7-mcp@latest"] | ||
|
|
||
| [mcp_servers.exa] |
There was a problem hiding this comment.
question (bug_risk): Validate that the mcp_servers.exa configuration using only a url matches the expected MCP server schema.
exa is configured only with a url, whereas the others use command/args. If the MCP runtime expects the same shape for all entries or special handling for HTTP servers, this could cause runtime issues or the server being ignored. Please confirm this format against the current config-schema.json and the Exa MCP docs.
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
There was a problem hiding this comment.
4 issues found across 11 files
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name=".agents/skills/agent-skills/SKILL.md">
<violation number="1" location=".agents/skills/agent-skills/SKILL.md:1">
P2: Remove the outer Markdown code fences so the document renders as Markdown instead of a single code block.</violation>
</file>
<file name=".codex/AGENTS.md">
<violation number="1" location=".codex/AGENTS.md:26">
P3: This sentence references workflow files that do not exist, which contradicts the previous line and can mislead users.</violation>
</file>
<file name=".codex/config.toml">
<violation number="1" location=".codex/config.toml:14">
P2: Pin MCP package versions in the baseline config instead of using floating `@latest`/unversioned `npx` installs.</violation>
</file>
<file name=".claude/skills/agent-skills/SKILL.md">
<violation number="1" location=".claude/skills/agent-skills/SKILL.md:1">
P2: The file is wrapped in a fenced code block, so the entire SKILL.md renders as literal code instead of documentation.</violation>
</file>
Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.
| @@ -0,0 +1,88 @@ | |||
| ```markdown | |||
There was a problem hiding this comment.
P2: Remove the outer Markdown code fences so the document renders as Markdown instead of a single code block.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At .agents/skills/agent-skills/SKILL.md, line 1:
<comment>Remove the outer Markdown code fences so the document renders as Markdown instead of a single code block.</comment>
<file context>
@@ -0,0 +1,88 @@
+```markdown
+# agent-skills Development Patterns
+
</file context>
|
|
||
| [mcp_servers.context7] | ||
| command = "npx" | ||
| args = ["-y", "@upstash/context7-mcp@latest"] |
There was a problem hiding this comment.
P2: Pin MCP package versions in the baseline config instead of using floating @latest/unversioned npx installs.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At .codex/config.toml, line 14:
<comment>Pin MCP package versions in the baseline config instead of using floating `@latest`/unversioned `npx` installs.</comment>
<file context>
@@ -0,0 +1,48 @@
+
+[mcp_servers.context7]
+command = "npx"
+args = ["-y", "@upstash/context7-mcp@latest"]
+
+[mcp_servers.exa]
</file context>
| @@ -0,0 +1,88 @@ | |||
| ```markdown | |||
There was a problem hiding this comment.
P2: The file is wrapped in a fenced code block, so the entire SKILL.md renders as literal code instead of documentation.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At .claude/skills/agent-skills/SKILL.md, line 1:
<comment>The file is wrapped in a fenced code block, so the entire SKILL.md renders as literal code instead of documentation.</comment>
<file context>
@@ -0,0 +1,88 @@
+```markdown
+# agent-skills Development Patterns
+
</file context>
|
|
||
| - No dedicated workflow command files were generated for this repo. | ||
|
|
||
| Use these workflow files as reusable task scaffolds when the detected repository workflows recur. No newline at end of file |
There was a problem hiding this comment.
P3: This sentence references workflow files that do not exist, which contradicts the previous line and can mislead users.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At .codex/AGENTS.md, line 26:
<comment>This sentence references workflow files that do not exist, which contradicts the previous line and can mislead users.</comment>
<file context>
@@ -0,0 +1,26 @@
+
+- No dedicated workflow command files were generated for this repo.
+
+Use these workflow files as reusable task scaffolds when the detected repository workflows recur.
\ No newline at end of file
</file context>
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
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.
Summary
Auto-generated ECC bundle from repository analysis.
What This Does
Merging this PR adds repo-local ECC artifacts for both Claude Code and Codex. The generated bundle captures repository patterns, Codex baseline config, and reusable workflow scaffolds derived from git history analysis.
Analysis Scope
Files
.claude/ecc-tools.json.claude/skills/agent-skills/SKILL.md.agents/skills/agent-skills/SKILL.md.agents/skills/agent-skills/agents/openai.yaml.claude/identity.json.codex/config.toml.codex/AGENTS.md.codex/agents/explorer.toml.codex/agents/reviewer.toml.codex/agents/docs-researcher.toml.claude/homunculus/instincts/inherited/agent-skills-instincts.yamlOptional: Continuous Learning (14 instincts)
This PR also includes instincts for the continuous-learning-v2 skill. These are optional and only useful if you use that skill.
Import after merging:
Review Checklist
ECC Tools | Everything Claude Code
Summary by Sourcery
Add repository-specific ECC configuration and skills for Claude and Codex, including multi-agent setups and learned development patterns.
New Features:
Enhancements:
Documentation:
Summary by cubic
Adds a repo-local ECC bundle for Claude Code and Codex, including a generated repo skill, Codex baseline with multi-agent roles, and a suggested
identity.json. This makes repo patterns and workflows available in-tool and standardizes local agent setup.New Features
.claude/skills/agent-skills/SKILL.mdand Codex copy at.agents/skills/agent-skills/SKILL.mdwith UI metadata at.agents/skills/agent-skills/agents/openai.yaml..codex/config.tomlenabling MCP servers:@modelcontextprotocol/server-github,@upstash/context7-mcp, Exa,@modelcontextprotocol/server-memory,@playwright/mcp,@modelcontextprotocol/server-sequential-thinking..codex/agents/*..claude/identity.jsonaligned to repo conventions..claude/homunculus/instincts/inherited/agent-skills-instincts.yaml.Migration
/instinct-import .claude/homunculus/instincts/inherited/agent-skills-instincts.yaml.Written for commit 7a725d8. Summary will update on new commits.