[github-actions]

Updated projects

Apache Airflow

Rules changed, ratings unchanged (AI-friendly with conditions)

• Added: Documentation: Allowed with conditions (Disclose AI use: State Gen-AI use in the PR description; Review AI output first: Review and understand what the tool produced before including it; Take full responsibility: Be prepared to explain and justify the Gen-AI use)
• Added: Pull request descriptions: Allowed with conditions (Disclose AI use: PR description must state that Gen-AI tools were used)

Confidence: medium

Needs a closer look:

• Dropped a quote not found verbatim in https://github.com/apache/airflow/blob/main/contributing-docs/05_pull_requests.rst: "Make sure to run все relevant static checks and tests locally to verify that the generated code works as intended and do"
• The sources point to more AI rules in https://www.apache.org/legal/generative-tooling.html, which is not a source yet: Apache Software Foundation Generative Tooling guidelines that Airflow requires Gen-AI users to follow.
• The sources point to more AI rules in ../COMMUNITY_ESCALATION.md, which is not a source yet: Community escalation process describing the further steps, including blocking and appeals, for repeated disregard of the Gen-AI guidelines.
• summaries: rated "unspecified" but has rules.

Bevy

Rules changed, ratings unchanged (AI-friendly with conditions)

• First detailed rule list: 19 rules.

Confidence: medium

Needs a closer look:

• The sources point to more AI rules in Bevy Code of Conduct, which is not a source yet: The AI Policy says the Code of Conduct continues to apply when AI-related topics are discussed.

curl

Area ratings changed, overall rating unchanged (AI-friendly)

• Added: Translating the software: Allowed (Disclose AI use (recommended): Consider mentioning you used an AI translation tool)
• Changed: AI-written reports: Discouraged → AI-written reports: Discouraged (Take full responsibility: Write the report yourself and explain the problem as you learned it)
• Changed: Security reports: Allowed with conditions (Disclose AI use: If AI was used to find the problem, reveal that fact in the report.; Review AI output first: Carefully verify AI findings yourself before reporting them.) → Security reports: Allowed with conditions (Disclose AI use: Reveal that an AI tool was used to find the problems; Review AI output first: Double-check the findings before reporting; Take full responsibility: Write the report yourself and explain the problem as you learned it)
• Added: Pull request descriptions: Discouraged
• Added: Comments: Discouraged
• Issues & bug reports: Mostly friendly (Disclose, Human review) → Mostly friendly (Disclose, Accountable, Human review)

Confidence: medium

Gentoo

Rules changed, ratings unchanged (Not AI-friendly)

• First detailed rule list: 23 rules.

Confidence: high

scikit-image

Rating changed (Not AI-friendly → AI-friendly with conditions)

• First detailed rule list: 11 rules.
• Code contributions: Unfriendly (Disclose, Accountable, License-safe, Minor help only) → Mostly friendly (Disclose, Accountable, License-safe, Limited scope)
• Issues & bug reports: Unfriendly (Minor help only) → No policy
• AI-written text: Unfriendly (Minor help only) → Unfriendly

Confidence: high

stb

Rules changed, ratings unchanged (Not AI-friendly)

• First detailed rule list: 18 rules.

Confidence: high

Traefik

Area ratings changed, overall rating unchanged (AI-friendly with conditions)

• First detailed rule list: 11 rules.
• Issues & bug reports: Mostly friendly (Disclose, Human review) → No policy
• AI-written text: Unfriendly (Minor help only) → Friendly (Accountable)

Confidence: high

Needs a closer look:

• The sources point to more AI rules in CONTRIBUTING.md, which is not a source yet: AGENTS.md defers to CONTRIBUTING.md and docs/content/contributing/ for everything not covered in the guide.
• The sources point to more AI rules in docs/content/contributing/submitting-security-issues.md, which is not a source yet: Security reports must follow this page's threat model and submission requirements, including the AI-disclosure rule.
• The sources point to more AI rules in docs/content/contributing/security-decisions.md, which is not a source yet: Lists settled security positions; reports matching a settled position will be closed.
• issues: rated "unspecified" but has rules.

Turso

Area ratings changed, overall rating unchanged (AI-friendly)

• First detailed rule list: 7 rules.
• AI-written text: Friendly (Human review) → No policy

Confidence: high

Needs a closer look:

• summaries: rated "unspecified" but has rules.

uutils coreutils

Rating changed (AI-friendly → AI-friendly with conditions)

• First detailed rule list: 8 rules.
• Code contributions: Friendly (Accountable, Human review, License-safe) → Mostly friendly (Accountable, Human review, License-safe, Limited scope)

Confidence: high

Vim

Area ratings changed, overall rating unchanged (AI-friendly with conditions)

• First detailed rule list: 10 rules.
• Code contributions: Mostly friendly (Disclose) → Mostly friendly (Disclose, Accountable, License-safe, Limited scope)
• AI-written text: No policy → Unfriendly (Minor help only)

Confidence: high

Needs a closer look:

• The sources point to more AI rules in AGENTS.md, which is not a source yet: Contains information and instructions for AI coding agents.
• The sources point to more AI rules in CODE_OF_CONDUCT.md, which is not a source yet: Covers the use of AI in discussions.

vLLM

Area ratings changed, overall rating unchanged (AI-friendly with conditions)

• First detailed rule list: 11 rules.
• Code contributions: Mostly friendly (Disclose, Accountable, Human review, No agents) → Mostly friendly (Disclose, Accountable, Human review)

Confidence: high

Needs a closer look:

• The sources point to more AI rules in tests/AGENTS.md, which is not a source yet: Domain-specific agent instructions for the test suite, referenced from AGENTS.md.
• The sources point to more AI rules in rust/AGENTS.md, which is not a source yet: Domain-specific agent instructions for the Rust frontend.
• The sources point to more AI rules in docs/contributing/editing-agent-instructions.md, which is not a source yet: Rules for modifying AGENTS.md and the domain-specific guides it references.
• The sources point to more AI rules in .agents/skills/pr-checklist/SKILL.md, which is not a source yet: Agent skill checklist that must be run and addressed before opening a PR or requesting re-review.

WordPress

Area ratings changed, overall rating unchanged (AI-friendly with conditions)

• First detailed rule list: 18 rules.
• Issues & bug reports: Friendly (Human review) → Mostly friendly (Disclose, Human review)
• AI-written text: Friendly (Human review) → Mostly friendly (Disclose, Human review)

Confidence: high

Needs a closer look:

• Statement 22 is not used by any rule: "If a tool seems to output code that looks suspiciously like a well-known library or non-GPL-compatible project, do not s"
• Statement 25 is not used by any rule: "- Run the test suites that are relevant to your change and note the commands you used in the PR description."

yt-dlp

Rules changed, ratings unchanged (Not AI-friendly)

• First detailed rule list: 24 rules.

Confidence: high

Review each change against its snapshot diff before merging.