Repository navigation
chore(aiinfo): release 2.11.18 - #228
Merged
Merged
Conversation
Contributor
There was a problem hiding this comment.
Copilot review overview
🟡 Changes recommended
Career Agent receives unrelated Skill Extraction risk text, and the new feature’s changelog entry is incomplete.
Review effort: Balanced
Findings: 2
Open (3)
What changed in this PR
Refreshes AiInfo data and releases version 2.11.18 with Career Skill Extraction.
Changes:
- Adds and exports Skill Extraction metadata.
- Updates Career Agent risk data.
- Refreshes generated bundles, cache, declarations, and changelog.
| File | Description |
|---|---|
packages/aiinfo/src/index.mjs |
Updates generated ESM bundle. |
packages/aiinfo/src/index.cjs |
Updates generated CommonJS bundle. |
packages/aiinfo/src/index.d.mts |
Declares the new ESM export. |
packages/aiinfo/src/index.d.cts |
Declares the new CommonJS export. |
packages/aiinfo/package.json |
Bumps version to 2.11.18. |
packages/aiinfo/node/index.ts |
Exports Skill Extraction. |
packages/aiinfo/node/components/careerskillsextraction/index.ts |
Defines Skill Extraction metadata. |
packages/aiinfo/node/components/careerlearnerchat/index.ts |
Reorders generated type imports. |
packages/aiinfo/node/components/careeragent/index.ts |
Updates expected-risk metadata. |
packages/aiinfo/Changelog.md |
Records refreshed data changes. |
packages/aiinfo/cache/checksum.json |
Updates dataset checksums. |
packages/aiinfo/cache/cache.csv |
Refreshes source spreadsheet data. |
💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.
| 046f8d336ed6e2bcb9a0b83642d4fe4bb4c7810d9b46d1414c9c1ffe91fc1824 | ||
| ``` | ||
|
|
||
| ### careerskillsextraction |
| careerAgent,2026.09.30,Learning Provider Chat Experience,"Generate courses, assignments, pages. Manipulates content within human-in-the-loop experience. Ability to create/update courses and learning objects within an account/course.",Claude 4.5 Sonnet,Anthropic Claude models are provided via Instructure's in-house AI Platform.,No,Course,,,Logs data,"Request, response for service","Virginia, Oregon, Sydney, Montreal, Dublin, Frankfurt",,Not Exposed,,No,Yes,"When chat experience interprets the user's desire to create content, it surfaces a confirmation modal which allows the user to modify the generated content prior to creation","Service strips html content to ensure it doesn't violate security requirements. Content is generated via user scope, so user can't create content in instances where that permission is not available",Proposed content may contain inaccuracies,Shorten loop for authoring content within a course,Canvas Career,2,, | ||
| careerSkillsExtraction,2026.10.01,Skill Extraction,Use AI to automatically identify skills in learning content to create a clearer picture of learner capabilities.,Claude Haiku 4.5,Anthropic Claude models are provided via Instructure's in-house AI Platform.,No,Course,Canvas learning objects on the account,The model does not retain any data,Logs data,Structured logs via logging service. Prompts and responses are not logged.,Global,,Not Exposed,Instructor-authored course content only,No,Yes,"Review is post-hoc curation, not gated approval. If nothing is reviewed, extracted skills remain aligned.",,"Needs human review to validate suggested skills before relying on them | ||
| Prompt injection via assignment HTML (no sanitization before Cedar).",Assist learning providers in creating a skill taxonomy for thier learning content,Canvas Career,2,, | ||
| careerAgent,2026.09.30,Learning Provider Chat Experience,"Generate courses, assignments, pages. Manipulates content within human-in-the-loop experience. Ability to create/update courses and learning objects within an account/course.",Claude 4.5 Sonnet,Anthropic Claude models are provided via Instructure's in-house AI Platform.,No,Course,,,Logs data,"Request, response for service","Virginia, Oregon, Sydney, Montreal, Dublin, Frankfurt",,Not Exposed,,No,Yes,"When chat experience interprets the user's desire to create content, it surfaces a confirmation modal which allows the user to modify the generated content prior to creation","Service strips html content to ensure it doesn't violate security requirements. Content is generated via user scope, so user can't create content in instances where that permission is not available",Needs human review to validate suggested skills before relying on them,Shorten loop for authoring content within a course,Canvas Career,2,, |
| careerKeyTakeaways,2026.09.30,Key Takeaways,"Gives a short summary of the key takeaways from a selected passage, a page, or a file.",Haiku 4.5,Anthropic Claude models are provided via Instructure's in-house AI Platform.,No,Course,"The selected text, page, or file.",,Logs data,"Logs request metadata only: account, user ID, feature, and duration. Journey doesn't log prompts or responses.",Global,,Not Exposed,The model doesn't get names or emails. Pseudonymous IDs go to the AI gateway only.,No,Yes,Learners read the output. Nothing is saved.,Canvas permissions limit what content the model sees. Rate limits apply.,Takeaways may leave out or misstate content. Verify against the source.,Help learners spot the main points of course material.,Canvas Career,2,, | ||
| careerAgent,2026.09.30,Learning Provider Chat Experience,"Generate courses, assignments, pages. Manipulates content within human-in-the-loop experience. Ability to create/update courses and learning objects within an account/course.",Claude 4.5 Sonnet,Anthropic Claude models are provided via Instructure's in-house AI Platform.,No,Course,,,Logs data,"Request, response for service","Virginia, Oregon, Sydney, Montreal, Dublin, Frankfurt",,Not Exposed,,No,Yes,"When chat experience interprets the user's desire to create content, it surfaces a confirmation modal which allows the user to modify the generated content prior to creation","Service strips html content to ensure it doesn't violate security requirements. Content is generated via user scope, so user can't create content in instances where that permission is not available",Proposed content may contain inaccuracies,Shorten loop for authoring content within a course,Canvas Career,2,, | ||
| careerSkillsExtraction,2026.10.01,Skill Extraction,Use AI to automatically identify skills in learning content to create a clearer picture of learner capabilities.,Claude Haiku 4.5,Anthropic Claude models are provided via Instructure's in-house AI Platform.,No,Course,Canvas learning objects on the account,The model does not retain any data,Logs data,Structured logs via logging service. Prompts and responses are not logged.,Global,,Not Exposed,Instructor-authored course content only,No,Yes,"Review is post-hoc curation, not gated approval. If nothing is reviewed, extracted skills remain aligned.",,"Needs human review to validate suggested skills before relying on them | ||
| Prompt injection via assignment HTML (no sanitization before Cedar).",Assist learning providers in creating a skill taxonomy for thier learning content,Canvas Career,2,, |
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
Validation
After merge, tag automation publishes aiinfo to npm. I will ensure nutritionfacts is pinned to the published version and its Pages deploy follows.