[DO NOT MERGE] CI mirror of #9892 to run credential-gated tests - #9916
Closed
nishantmonu51 wants to merge 3 commits into
Closed
nishantmonu51 wants to merge 3 commits into
nishantmonu51 wants to merge 3 commits into
Conversation
On every turn, the analyst detects skill references in the prompt (<chat-reference>type="skill" skill="..."</chat-reference>) and pre-invokes load_skill once per distinct referenced skill that exists and applies to the analyst, before the model's turn. Unknown skills and skills for other agents are ignored; always-apply skills already pre-loaded on the first turn aren't loaded again. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Typing "/" in an analyst chat (project, dashboard, and embed) opens a picker listing the project's Skill resources that apply to the analyst, showing name and description. Picking one writes a <chat-reference>type="skill" skill="..."</chat-reference> tag into the prompt text -- the same tag the runtime resolves via load_skill -- and it renders as a chip both while typing and when the conversation is reloaded. A dedicated / button next to the existing @ button opens the same picker, adding a space first when the cursor sits right after a word. The "/" trigger is a second suggestion on the same Mention node as "@", so both share the parsing and serialization of chat-reference tags, the two pickers never open at once, and both follow Suggestion's default prefixes: a slash inside a word, as in a date or a path, stays text. The context picker takes the source of its options and whether several are picked, so the skills picker reuses it as a single-choice list. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A load_skill call whose result is an error, such as a skill that was not found while its file had an error, marked the skill as loaded, so a later reference to it was skipped and the model never got its body. Only calls with a successful result count now. Co-Authored-By: Claude Opus 5 (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
2 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.
Do not merge or review. Temporary mirror of #9892, opened only to run CI with repository secrets.
#9892 comes from a fork, so
pull_requestruns there receive no secrets. That makestest,e2e (web-admin)ande2e (web-integration)fail on missing credentials (RILL_RUNTIME_DRUID_TEST_DSN, the S3/GCS test keys, andRILL_BINARY_SAfor the devtool.env) regardless of the change. The same three checks fail identically on every open fork PR.This branch carries the same three commits, authored by @dfliess, so the credential-gated suites can actually run. Review and merge happen on #9892; this branch gets deleted once the run finishes.