Skip to content

docs(transformation): AI Transformation motion — Variant A (new top-level tab) - #19

Open
rs07-within wants to merge 9 commits into
mainfrom
applai-322-transformation-tab
Open

docs(transformation): AI Transformation motion — Variant A (new top-level tab)#19
rs07-within wants to merge 9 commits into
mainfrom
applai-322-transformation-tab

Conversation

@rs07-within

@rs07-within rs07-within commented Sep 10, 2026

Copy link
Copy Markdown

Linear

What this is

Expands the developer docs (developers.within.ai) with the AI Transformation motion — the Improve layer of Discover -> Structure -> Improve — per APPLAI-322. Built on the content from PR #11, brought forward onto the rebranded main (Klarity -> Within) and extended. All examples are synthetic; tool-call examples are verified against the live production Within MCP input schema (api.within.ai/mcp).

Two variants, opened as a pair

This is one of two PRs with identical page content; only docs/docs.json (the nav) differs, so the team can compare the two navigation shapes on their Mintlify previews and pick one. Do not merge until a variant is chosen and the APPLAI-368 consolidation is applied.

Pages

New: transformation/readouts.mdx (defensible, evidence-backed readouts + learnings); transformation/opportunity-loops.mdx (opportunity-scanning routines with a bounded clear/remediate/escalate loop).

Ported from PR #11 and revised: transformation.mdx (overview), objective, company-brain, use-case-selection, what-you-can-build (now carries the solution-space dimensions), where-you-build (Advisor vs MCP build location + the L3/L4 developer pull), non-technical-track (skill-build walkthrough), technical-track (ADK eight-step sequence), maintain-agents.

Enriched: guides/find-opportunities.mdx — a synthetic ranked sample output.

This variant: new top-level tab

Adds a new top-level AI Transformation tab holding the transformation pages as a Playbook group. Highest visibility, largest nav change.

Overlap with the proposed user-docs MCP page (APPLAI-368)

The Agent Platform team has a proposed user-facing page "Using the Within MCP" (docs.within.ai, PR #27). It and this draft are complementary — that page is the user "what/why" layer, this is the developer "how to build" layer. To consolidate before publish:

  • Advisor vs MCP boundary (overlap): where-you-build.mdx restates the Advisor / MCP / client boundary his page also covers. Make his page the canonical comparison; trim this to the build-location decision and link out once the page has a published URL.
  • Agent Builder / skills (overlap, clean): his page links to our /guides/agent-builder; build depth stays here.
  • Audit log (gap to fix here): his page states Within keeps a tool-call audit log; our security.mdx does not. Add it, pending confirmation with Agent Platform.
  • Uniquely ours (not in his page): transformation opportunities, readouts, opportunity loops, use-case selection, maintain agents.

Full section-by-section comparison and the consolidation proposal are on APPLAI-368.

🤖 Generated with Claude Code

Sibling variant (compare the other nav): #20

rs07-within and others added 9 commits September 10, 2026 17:52
Bring the docs/transformation pages from PR #11 (add-transformation-playbook)
forward onto current main, and sweep Klarity -> Within in prose so they match
the rebranded site. Nav wiring and the new pages (readouts, opportunity loops)
follow in later commits.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
New docs/transformation/readouts.mdx: how to produce a defensible, evidence-backed
readout (business case first, observed vs inferred, depth before compression,
interactive diagrams, pre- vs post-sale), with a synthetic prompt/output and a
learnings section. Drafted by GPT 5.6 via Codex from REQUIREMENTS-FOR-CODEX.md (W1);
verified against the brief's acceptance checks. Nav wiring follows in the nav commit.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Document a bounded recurring scan, add a synthetic ranked opportunity output, and make the readout evidence pull reproducible. The transformation overview now links the new evidence and readout paths.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Document the solution as dimensions instead of named tiers. Add the full customer-facing skill walkthrough because the existing track did not show the approval gates, package shape, or a concrete configuration.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Fold Derrick's Advisor and MCP positioning into the existing page. Separate Advisor, the read-only Within MCP, and client-owned actions, and correct ADK's role from a surface to a builder method.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace internal shorthand, promotional claims, and unsupported absolutes with explicit evidence and control guidance. Correct maintenance to match the read-only MCP contract: the client owns baselines and diffs, while Within supplies current process data.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add the hierarchy-node drill-down and name the boundary between the L3 process graph and L4 activity timelines so developers can choose the smallest evidence pull.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Remove two negative-parallelism phrases flagged in review without changing the verified MCP examples or page flow.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Create the top-level tab variant and remove the duplicated opportunity guide from How-to. Place readouts before use-case selection because they turn evidence into the decision brief, and place Where you build before the two implementation tracks.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
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