Skip to content

Support multiple tasks in the autowebcompat agent#6282

Draft
jgraham wants to merge 2 commits into
mozilla:masterfrom
jgraham:refactor_multiple_tasks
Draft

Support multiple tasks in the autowebcompat agent#6282
jgraham wants to merge 2 commits into
mozilla:masterfrom
jgraham:refactor_multiple_tasks

Conversation

@jgraham

@jgraham jgraham commented Jul 3, 2026

Copy link
Copy Markdown
Member

No description provided.

jgraham added 2 commits July 2, 2026 12:33
To pass inputs ot agents in production, upload a file to a GCS bucket,
and pass a URL to that file into the agent environment.

This enables having inputs which don't fit inside the 32kB limits of
enironment variables for GCS.

For local runs using docker, data is still pased via environment variables.
Currently the agent does two things: checks in nightly and checks with
Chrome Mask enabled. These are done in a single Firefox session and a
single invocation of the LLM. However this approach won't scale; if we
want to check on multiple browsers, multiple Firefox versions,
etc. putting all the logic into the LLM prompt will be fragile and
hard to debug.

Instead create a base Task class that knows how to invoke the LLM and
get results out. Then create a specific subclass for each kind of
operation we're interested in, and write some business logic to
deterministically wire them together according to our needs.
@jgraham

jgraham commented Jul 3, 2026

Copy link
Copy Markdown
Member Author

CC @ksy36

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant