Skip to content

docs: close Coder Agents coverage gaps and align nav references#24971

Merged
mattvollmer merged 13 commits into
mainfrom
matt/docs-agents-coverage-gaps
May 5, 2026
Merged

docs: close Coder Agents coverage gaps and align nav references#24971
mattvollmer merged 13 commits into
mainfrom
matt/docs-agents-coverage-gaps

Conversation

@mattvollmer
Copy link
Copy Markdown
Contributor

@mattvollmer mattvollmer commented May 5, 2026

Closes coverage gaps in docs/ai-coder/agents/ and aligns nav references with the current UI (post #24574 Behavior split, post #24644 Insights removal).

Content fixes:

  • Replace site-wide coder users edit-roles flow with org-scoped agents-access role (per migration 000475). CLI examples now preserve existing org roles since edit-roles overwrites the full set.
  • Correct computer-use claim: supports Anthropic and OpenAI providers, configured under the Virtual desktop experiment.
  • New platform-controls/experiments.md covering Virtual desktop, Advisor, and Chat debug logging (each as: what, how to enable, API). Includes the Debug tab in the chat right panel.
  • Trim models.md "Model overrides" to essentials: two layers (admin subagent, user personal), contexts table, resolution order, API pointer.
  • Remove retired platform-controls/pr-insights.md (page + manifest + cross-links).

Nav cleanup:

  • Admin-only tabs use the full Agents > Settings > Manage Agents > <Tab> path; user-side tabs keep Agents > Settings > <Tab>.
  • Replace stale "Behavior" references with Instructions / Lifecycle / Experiments to match the current sidebar.
  • Replace references to the removed top-bar Admin dialog with the Settings sidebar.
Decision log
  • Experimental features were originally drafted as a standalone Advisor page plus inline sections in platform-controls/index.md. Consolidated into one experiments.md since no individual feature warrants a full page yet and parallel short sections are easier to scan.
  • Reviewer feedback on early drafts: drop the inline experiments list from index.md (avoid drift), drop the "users created before this role was introduced" note (handled transparently by migration 000475), specify the full nav path for per-model pricing, link the type=computer_use row in architecture.md to the Experiments page.
  • CLI bulk-grant script previously called edit-roles <user> agents-access. That replaces the user's full org role set, so the script would silently strip organization-admin, organization-template-admin, etc. Rewrote to read each user's current roles, append agents-access, dedupe, and write the union back.

PR generated with Coder Agents.

Fix nav references that no longer match the Settings sidebar in the
Agents page after the Behavior page split (#24574) and Insights nav
removal (#24644).

- Plan mode instructions: Behavior -> Instructions
- Workspace autostop fallback: Behavior -> Lifecycle
- Data retention: Behavior -> Lifecycle
- System prompt: add explicit Settings > Instructions path
- Add provider, Add model, system prompt config: replace removed
  "Admin" top-bar dialog with Settings panel
- Personal model overrides: drop redundant 'Manage Agents' hop and
  match the convention used by sibling docs
- PR Insights: note that the page is no longer linked from the
  sidebar and is reachable only by direct URL or API
- Image alt text and 'Admin panel' phrasing updated to match the
  current Settings panel name
@github-actions
Copy link
Copy Markdown

github-actions Bot commented May 5, 2026

Docs preview

📖 View docs preview for docs/ai-coder/agents/architecture.md

Remove the PR Insights page and references to it from sibling docs
and the manifest. The dashboard at /agents/settings/insights is no
longer linked from any UI surface and the page is being retired.
@mattvollmer mattvollmer changed the title docs(ai-coder/agents): close coverage gaps and align nav references docs(docs/ai-coder/agents): close coverage gaps and align nav references May 5, 2026
@mattvollmer mattvollmer changed the title docs(docs/ai-coder/agents): close coverage gaps and align nav references docs: close Coder Agents coverage gaps and align nav references May 5, 2026
…l features into one page

Replace the standalone Advisor page and the two experimental sections
in index.md with a single Experiments page. The new page follows the
Experiments tab in Settings and covers Virtual desktop, Advisor, and
Chat debug logging in short, parallel sections.

The index page now has one Experiments pointer instead of three
sections inline, matching how sibling experimental docs are surfaced
elsewhere.
…y tabs

Insert the 'Manage Agents' sub-panel into every nav reference that points
to an admin-only Settings tab. The Settings sidebar puts Agents, Providers,
Models, MCP Servers, Templates, Spend, Instructions, Experiments, and
Lifecycle behind a 'Manage Agents' drilldown, so the short form
'Settings > <admin-tab>' was understating one click.

User-side tabs (General, Agents personal overrides, Compaction, Secrets/
API Keys) keep the short form since they sit at the top of the Settings
panel.
Comment thread docs/ai-coder/agents/platform-controls/index.md Outdated
… panel

The right panel on the Agents page shows a Debug tab when debug logging
is active for the chat. Earlier text claimed there was no dashboard
viewer; correct that and keep the API reference for programmatic use.
Comment thread docs/ai-coder/agents/architecture.md Outdated
Comment thread docs/ai-coder/agents/getting-started.md Outdated
Comment thread docs/ai-coder/agents/getting-started.md Outdated
Comment thread docs/ai-coder/agents/getting-started.md Outdated
Comment thread docs/ai-coder/agents/models.md Outdated
…d and experiments

Addresses review comments on PR #24971:
- getting-started.md: drop the upgrade-backfill note (it's noise for new
  installs; the migration handles it transparently).
- getting-started.md: rework the CLI section. The `edit-roles` command
  replaces the user's full org role set, not append to it, so the prior
  examples would silently drop other org roles. Replace with examples
  that fetch the current roles, add `agents-access`, and write the
  union back.
- getting-started.md: spell out the navigation path for per-model token
  pricing ("Manage Agents > Models").
- platform-controls/index.md: drop the inline list of current
  experiments. The "See Experiments" link is the source of truth, so
  this section doesn't need to be kept in sync.
- architecture.md: link the `spawn_agent` (`type=computer_use`) row to
  the virtual desktop experiment so users can see how to enable it.
…rage-gaps

# Conflicts:
#	docs/ai-coder/agents/architecture.md
#	docs/ai-coder/agents/platform-controls/index.md
The previous section was a near-complete API reference that mostly
repeated what the experimental chat configuration API already documents.
For the getting-started narrative, keep:

- The two override layers (admin subagent, user personal) and where to
  find them in the UI.
- The configurable contexts in one table.
- The resolution order in three lines.
- A pointer to the experimental config API.

Drop the per-endpoint URLs, request/response shapes, error codes, and
mode-resolution mechanics. Anyone wiring this up programmatically should
read the API definition rather than the docs page.
@mattvollmer mattvollmer marked this pull request as ready for review May 5, 2026 16:30
@mattvollmer mattvollmer requested review from david-fraley and ibetitsmike and removed request for ibetitsmike May 5, 2026 16:30
Copy link
Copy Markdown
Collaborator

@david-fraley david-fraley left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

stamping

@mattvollmer mattvollmer merged commit e189f73 into main May 5, 2026
35 checks passed
@mattvollmer mattvollmer deleted the matt/docs-agents-coverage-gaps branch May 5, 2026 16:40
@github-actions github-actions Bot locked and limited conversation to collaborators May 5, 2026
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants