Skip to content

docs(TSP-1446): document Inventor Routines feature - #815

Open
claude[bot] wants to merge 3 commits into
mainfrom
docs/TSP-1446
Open

claude[bot] wants to merge 3 commits into
mainfrom
docs/TSP-1446

Conversation

@claude

@claude claude Bot commented Sep 15, 2026

Copy link
Copy Markdown

Summary

  • Replaced the single "Scheduled check-ins" accordion in the Manage performance tab with two distinct accordions: Routines (recurring scheduled instructions) and Check-ins (one-time wake-ups)
  • Routines accordion documents: schedule options (daily/weekly/monthly/annually), per-creator permission model, self-contained instruction requirement, time-of-day picker, and the unified schedule management modal
  • Check-ins accordion updated to clarify its one-time nature and cross-link to routines
  • Added three new FAQ entries covering the routine vs check-in distinction, edit/delete permissions, and how to write routine instructions
  • Feature gate note included in the Routines accordion

Linear issue

https://linear.app/relevance/issue/TSP-1446/

Test plan

  • Verify both new accordions render correctly in Mintlify preview
  • Confirm feature gate Note is visible in Routines accordion
  • Check all internal anchor links resolve (#alerts)
  • Verify FAQ entries appear in the AccordionGroup at the bottom of the page

🤖 Generated with Claude Code

Add documentation for routines (recurring scheduled instructions) and
update the check-ins accordion to reflect its role as a one-time
wake-up. Covers schedule options, permission model, the unified
schedule management modal, and three new FAQ entries.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
@claude claude Bot added the docs-drafter Documentation drafted by Claude label Sep 15, 2026
@mintlify

mintlify Bot commented Sep 15, 2026

Copy link
Copy Markdown
Contributor

Preview deployment for your docs. Learn more about Mintlify Previews.

Project Status Preview Updated
relevanceai 🟢 Ready View Preview Sep 15, 2026, 3:07 AM

💡 Tip: Enable Automations to automatically generate PRs for you.

@linear

linear Bot commented Sep 15, 2026

Copy link
Copy Markdown

TSP-1446

Adds documentation for the new Routines feature (PRs #17993-17995):
- New Routines accordion: recurring scheduled instructions, schedule options (daily/weekly/monthly/annually), self-contained instruction requirements, and the creator/team-member permissions model
- Updated Scheduled check-ins accordion: clarifies one-time, conversation-specific nature and how Invent suggests check-ins for one-time follow-ups
- New Schedule management accordion: documents the unified modal showing check-ins and routines with a granular time picker
- Two new FAQ entries covering check-in vs routine distinction and routine editing permissions

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
@github-actions

Copy link
Copy Markdown
Contributor

🎯 Vibe check

Reviewed: 1 file (1 with issues, 0 clean)

Scores

Dimension Score What's holding it back
🟢 Consistency 9/10 **Powerful** on line 151 is a banned word. It's used as a UI label name, but should still be verified — if the product literally shows "Powerful" in the dropdown, documenting it accurately is fine; if it's paraphrase, reword it.
🟡 Technical clarity 8/10 Two anchor links (#alerts) target an accordion nested inside a tab — this pattern may not resolve reliably in Mintlify.
🟢 Non-technical clarity 9/10 Clear definition up front, strong use of CardGroup for capabilities, good "MCP vs Invent" comparison section for confused newcomers.
🟡 Structure 7/10 Four FAQ entries duplicate their corresponding enterprise security accordion content verbatim. Page ends after FAQs with no closing CTA.

Score key: 🟢 9–10, 🟡 6–8, 🔴 1–5.

Overall vibe: This is a well-written, comprehensive hub page — precise language, correct component usage, solid coverage of access controls, features, and Evals integration. The main structural debt is that the FAQ section copy-pastes four full paragraphs from the enterprise security accordions immediately above; those FAQs should link to the section instead. The page also needs a closing CTA to give readers somewhere to go after they've finished reading.

🔧 Issues (3)
  • build/invent/invent.mdx:151**Powerful** is a banned word. If this is the exact label the product shows in the model dropdown, document it accurately and keep it; if it's prose characterizing the model, reword it (e.g., "high-capability").
  • build/invent/invent.mdx:214[Alerts](/build/invent/invent#alerts) links to an accordion title inside a Tab component. Mintlify anchors are generated from headings, not accordion titles nested in tabs — this link will likely scroll to the top of the page rather than the Alerts accordion. Verify it resolves as expected; if not, convert the Alerts accordion into a heading or remove the anchor.
  • build/invent/invent.mdx:311 — Same #alerts anchor issue as line 214.
🏗️ Page structure (3)
  • build/invent/invent.mdx:349–363 — Four FAQ entries ("Who can use Invent?", "How does project isolation work for Enterprise customers?", "Which integrations can Invent access in an Enterprise project?", "Where can Enterprise Invent data be sent?") are verbatim copies of content already in the Availability section (lines 44–46) and Enterprise security controls accordions (lines 91–101). Replace the duplicated FAQ bodies with a one-sentence summary and a link to the relevant section, e.g. "See Enterprise security controls above."
  • build/invent/invent.mdx — No closing CTA. This is a concept page and ends abruptly after the FAQ accordion. Add a short ## What's next? or a single <Card> pointing readers onward — e.g. Version history (mentioned inline but never highlighted as a follow-on) and Introduction to Evals.
  • build/invent/invent.mdx:110–112 — The credit <Warning> sits between the "How to use Invent" video and the "Invent features" section. It reads as disconnected from both. Consider moving it immediately before the video (so readers see the cost caveat before they watch the demo) or collapsing it into the "Model selection and credit usage" accordion where cost is already explained.
⚠️ Contradictions (0)

No contradictions found between this file and the context pages read (version-history.mdx, evals/introduction.mdx). The restore-to-draft behavior and (restored) version naming are consistent across both files.

🔋 Credit usage
Item Count
Files reviewed 1
Context pages read 2
Total lines processed ~580

Files read: build/invent/invent.mdx (389 lines), build/agents/build-your-agent/version-history.mdx (89 lines), build/agents/evals/introduction.mdx (105 lines)

…tion

Elevate the ownership/permissions model to a prominent callout, document
the unified schedule management modal in more detail, and add cross-links
to User Level Authentication, RBAC, Integrations, and Agent Scheduled
Triggers pages.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
@github-actions

Copy link
Copy Markdown
Contributor

🎯 Vibe check

Reviewed: 1 file (1 with issues, 0 clean)

Scores

Dimension Score What's holding it back
🟢 Consistency 8/10 One capitalization miss (toolsTools on line 336); "The Inventor" used once instead of "Invent" (line 213).
🟡 Technical clarity 7/10 Two anchor links point to an <Accordion> title rather than a heading — Mintlify may not resolve them. "Configure an alarm" tab opens with auto-diagnosis instructions before telling the reader what an alarm is.
🟢 Non-technical clarity 8/10 "Forward deployed agent" (lines 4, 11) is jargon left undefined, though the sentences that follow do explain the function. Everything else is plain-English and well-paced.
🟡 Structure 7/10 No closing CTA on a concept page that ends with an FAQ. Alarm tab has inside-out ordering.

Score key: 🟢 9–10, 🟡 6–8, 🔴 1–5.

Overall vibe: This is a well-organized, comprehensive reference page — good use of Tabs, Steps, and AccordionGroups, zero banned words, clean sentence case throughout, no marketing fluff. The main issues are two self-referential anchor links that may silently break (the target is an Accordion, not a heading), one naming inconsistency, and a missing CTA that would give readers somewhere to go after reading.

🔧 Issues (4)
  • build/invent/invent.mdx:336 — "Call your existing Relevance AI tools from external clients" → Tools (capital T). These are Relevance AI product features, same as the capitalized Tools used throughout the rest of the page.
  • build/invent/invent.mdx:213 — "The Inventor will also suggest a check-in…" → "Invent will also suggest a check-in…". "Inventor" appears only as part of the feature name "Inventor Routines"; using "The Inventor" as a standalone synonym for the product is inconsistent with every other sentence on the page.
  • build/invent/invent.mdx:218[Alerts](/build/invent/invent#alerts) links to an <Accordion title="Alerts">, not a heading. CLAUDE.md requires anchor targets to be proper headings (# through ######). If Mintlify doesn't generate an id for accordion titles, this resolves to the top of the page. Verify in preview, or promote "Alerts" to a #### heading inside the Manage performance tab.
  • build/invent/invent.mdx:315[Invent alert channels](/build/invent/invent#alerts) — same anchor issue as line 218.
🧩 Component suggestions (1)
  • build/invent/invent.mdx:313–322 — The "Configure an alarm" tab opens with a heading ("Turn on auto-diagnosis") and explains how to enable that feature, then drops the definition of what an alarm actually is at the very end: "Alarms evaluate the scores collected by a Performance dashboard and notify you when an overall score or individual Check crosses the condition you set." That sentence should be the first thing in the tab body — readers need to know what they're configuring before they see the auto-diagnosis toggle. Move it above the #### Turn on auto-diagnosis heading.
🏗️ Page structure (1)
  • build/invent/invent.mdx — Concept page with no closing CTA. The FAQ is the last thing on the page and leaves readers with no next step. Add a ## What's next? section (or a <CardGroup>) pointing to /build/agents/evals/introduction ("Start measuring Agent performance with Evals") and /build/agents/build-your-agent/overview ("Build your first Agent") — both are naturally where a reader who just learned about Invent would go next.
🔋 Credit usage
Item Count
Files reviewed 1
Context pages read 2
Total lines processed ~860

Files read: build/invent/invent.mdx (393 lines), build/agents/evals/introduction.mdx (105 lines), integrations/mcp/mcp-server.mdx (362 lines)

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

Labels

docs-drafter Documentation drafted by Claude

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant