Repository navigation
AGENTS.md: clarify signed-off and co-developed-by guidance - #11333
Conversation
| * AI agents MUST NOT add Signed-off-by tags. The human submitter is responsible for: | ||
| * Reviewing all AI-generated code | ||
| * Ensuring compliance with licensing requirements | ||
| * Adding their own Signed-off-by tag to certify the DCO | ||
| * Taking full responsibility for the contribution | ||
| * Co-developed-by usage guidelines follow upstream Linux conventions. Specifically | ||
| * Co-developed-by: states that the patch was co-created by multiple developers; it is used to give attribution to co-authors (in addition to the author attributed by the From: tag) when several people work on a single patch. Since Co-developed-by: denotes authorship, every Co-developed-by: must be immediately followed by a Signed-off-by: of the associated co-author. |
There was a problem hiding this comment.
We should just say what the agent needs todo here rather what the human should do (the human R&Rs would be in the README.md).
There was a problem hiding this comment.
@lgirdwood But I see lot of PRs where agents are adding these. AGENT.md seems the place to ensure this stops happening..
There was a problem hiding this comment.
yes - its fine to tell what the agent to do here. But we are also telling the human what to do here too, i.e.
The human submitter is responsible for:
* Reviewing all AI-generated code
* Ensuring compliance with licensing requirements
* Adding their own Signed-off-by tag to certify the DCO
* Taking full responsibility for the contribution
This is not used by the agent.
We need to make this agent only. i.e. tell the agent what to do directly here (not via reference to a large file on the internet) i.e.
AI agents MUST NOT add Signed-off-by tags but can add Assisted-by: tags where code has been developed by the agent.
One other thing is we should soft link this for CLAUDE.md too.
There was a problem hiding this comment.
@lgirdwood Let me add. Couple of notes:
I'd like to drop the assisted-by. This is what Linux upstream did in 7.3.0 as well.
CLAUDE.md -> we already have this, it points to AGENTS.md.
There was a problem hiding this comment.
Warning
Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.
Copilot review overview
1 open issue
What changed in this PR
Updates AGENTS.md to clarify Developer Certificate of Origin (DCO) responsibilities and attribution conventions, aligning guidance with upstream Linux kernel wording.
Changes:
- Explicitly prohibits AI agents from adding
Signed-off-bytags and clarifies human submitter responsibilities. - Adds guidance on
Co-developed-byusage and its required pairing with a correspondingSigned-off-by.
| File | Description |
|---|---|
| AGENTS.md | Clarifies sign-off/DCO ownership and Co-developed-by conventions to match upstream Linux guidance. |
Review effort: Lite
| * Ensuring compliance with licensing requirements | ||
| * Adding their own Signed-off-by tag to certify the DCO | ||
| * Taking full responsibility for the contribution | ||
| * Co-developed-by usage guidelines follow upstream Linux conventions. Specifically |
PR 11333: test resultsRun date: 2026-10-22 14:55 UTC Tested commit: 1e7ab9222b8eec05b5984e03d6110f08bcadf57d
|
Add explicit note that agents must not add signed-off-by lines. Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Add human-facing responsibilities for AI-assisted contributions. This complements AGENTS.md which governs agent behaviour, while README.md is the natural place for guidance aimed at human submitters. Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
36428b1 to
1e7ab92
Compare
|
V2:
|
Align SOF policy to upstream Linux kernel and specifically the wording adopted in Linux 7.3.0.
Link: https://docs.kernel.org/process/submitting-patches.html
Link: https://docs.kernel.org/process/coding-assistants.html