Add per-agent MCP toolset configuration to agent setup - #279
Merged
Conversation
Collaborator
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 149546516f
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
wpfleger96
force-pushed
the
sprout-ux/toolsets-ui
branch
from
April 9, 2026 20:12
1495465 to
b2cfbc5
Compare
Toolsets like canvas, channel_admin, dms, and workflow_admin could only be enabled by setting SPROUT_TOOLSETS in the shell environment before launching the desktop app — with no way to configure them per-agent in the UI. This made multi-agent setups where different agents need different tool scopes impractical. Adds an mcp_toolsets field to ManagedAgentRecord, CreateManagedAgentRequest, and UpdateManagedAgentRequest. The runtime reads this field and forwards it as SPROUT_TOOLSETS to the spawned MCP subprocess, overriding (or removing) any ambient env var. A comma-separated text input appears in the Advanced Setup section of the create-agent dialog, placed next to the MCP command field, with a help line listing the available toolset names.
Backticks in JSX strings render as visible characters, not as inline code formatting.
createManagedAgent() built the input object explicitly but omitted mcpToolsets, so the value collected in the dialog was silently dropped. Per-agent toolset configuration did nothing at creation time.
wpfleger96
force-pushed
the
sprout-ux/toolsets-ui
branch
from
April 10, 2026 16:34
b2cfbc5 to
b4c9ab2
Compare
Member
Author
tlongwell-block
added a commit
that referenced
this pull request
Apr 11, 2026
…ona-migration * origin/main: feat(desktop): add Pulse social notes surface (#296) Fix flaky desktop smoke tests (#294) Add agent lifecycle controls to channel members sidebar (#291) Update nest_agents.md tagging info (#292) feat: add Sprout nest — persistent agent workspace at ~/.sprout (#290) Fix auth and SSRF vulns (#261) Add per-agent MCP toolset configuration to agent setup (#279) feat(desktop): team & persona import/edit flows (#288) Remove menu item subtitles and fix persona card overflow (#289) feat: Phase 1 video upload support (Blossom-compliant-ish) (#285) Add inline subtitles to menu items and field descriptions (#276) Improve ephemeral channel affordances and hide archived sidebar rows (#286) Fix @mention search to use word-boundary prefix matching (#278) Allow bot owners to remove their agents from any channel (#284) [codex] Polish agent selectors and settings layout (#283) # Conflicts: # desktop/scripts/check-file-sizes.mjs
jeremylongshore
added a commit
to intent-solutions-io/buzz
that referenced
this pull request
Jul 29, 2026
…econcile the blueprint ledger to prod-up (#11) WHAT: Adds 000-docs/006-DR-STND-authoritative-naming-and-boundaries.md as the canonical source of truth for every Buzz asset name and the boundary between them (owner amendment 2026-07-29, FINAL). Reconciles the 001 blueprint completion ledger to reality and pins asset names in Phases 6/7. WHY: Two assets deliberately share the string "intent-ops-buzz" — a VPS host and a GitHub repository — and early planning blurred the code fork, the contribution/verification lab, and the live ops lane into one "Buzz repo". Conflating planes is how prod creds leak into a PR runner or an untrusted fork branch gets deployed. The owner amendment forbids building durable architecture docs (the contributor-lab architecture) until this naming record lands; this is Track B, ordered FIRST. Chose the fork 000-docs as its home (over the not-yet-created lab repo) because Track B requires no new repo and the record must land now to unblock Track C. WHERE (architecture): the fork is the code plane; this record governs the four-plane flow — code fork (submit upstream) → contributor-lab repository (execute tests) → Intent Eval Platform (decide) → intent-os ops/buzz (record) → intent-ops-buzz production host (run). Two separate gates: contribution (SUBMIT/HOLD/…) and deployment (ship/no_ship/…). CHANGES: - 006 (new): six-asset table, host-vs-repo disambiguation rule, four-plane flow diagram, two-gate table, cross-plane boundary rules. - 001 ledger: E2 → staging live + backups/restore proven; E2d → prod host built+deployed+verified pre-cutover (PRs block#279 block#280); E3 → in-flight with the remaining BLOCKING go-live gates enumerated; adds an ELab (contributor lab) row; E9 plugin named intent-ops-buzz-plugin + marked DEFERRED. - 001 phase tree: new Phase 6 (contributor-lab pointer, architecture graduates in Track C, not before); Phase 7 plugin repo name pinned. - 000-INDEX: 006 registered. VERIFIED: grep shows every intent-ops-buzz use disambiguated (production host vs repository) or fully org-qualified; content is public-safe (no hosts-as-addresses, IPs, keys, or member data). The contributor-lab holding draft stays untracked (it graduates into the lab repo in Track C). RISKS: none — additive-only IS-overlay docs; zero upstream-path edits. UNFINISHED: Track C (build the intent-ops-buzz repository) and the remaining Track A go-live gates. Bead: buzz-aet Co-authored-by: jeremylongshore <jeremylongshore@users.noreply.github.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
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.

This PR adds a
mcpToolsetsfield to the agent setup UI, replacing the globalSPROUT_TOOLSETSenv var as the only way to enable toolsets likecanvas,channel_admin, anddms.Previously, toolset selection required setting
SPROUT_TOOLSETSin the shell before launching the desktop app — no way to scope it per-agent, making multi-agent setups with different tool scopes impractical.mcp_toolsets: Option<String>toManagedAgentRecord,CreateManagedAgentRequest, andUpdateManagedAgentRequestintypes.rsruntime.rs, forwardmcp_toolsetsasSPROUT_TOOLSETSto the spawned MCP subprocess; remove it whenNoneto prevent inheriting a stale ambient env varmcpToolsetstext input in the Advanced Setup section ofCreateAgentDialogSections.tsxwith a help line listing all available toolset namesPost-merge tasks
mcpToolsetsintoEditAgentDialogafter Add edit dialog for managed agents with relay profile sync #277 lands (toolset field needs edit UI support)