feat(mcp): MCP UX improvement by jinhongkuan · Pull Request #201 · BicameralAI/bicameral-mcp

jinhongkuan · 2026-05-06T01:01:58Z

Summary

User-facing UX wins, in install order:

Setup wizard pre-approves bicameral MCP tools at the user level so post-meeting catch-up flows (pure-read ledger queries fired from a SessionStart brief) don't pop a permission prompt for every tool call. Implements §1 of the v0 productization plan. Writes only to ~/.claude/settings.json after showing the exact diff and requiring explicit y/N. Project .claude/settings.json is never touched.
Bash, Edit, Write, Read/Grep/Glob still prompt every time. The wizard auto-approves bicameral MCP tools only — bicameral.reset is deny-listed so destructive work always confirms. Tests guard the no-Bash invariant.
/bicameral-report-bug no longer shells out. Removed every Bash block (env probing, python3 -c URL construction, open/xdg-open/start). The skill now derives env from agent context + Read and prints a clickable URL — no shell-permission friction in the bug-report flow.
Skill telemetry no longer collects rationale. It was descriptive freeform text that risked leaking session context into PostHog without measurable analytic value.
extract_symbols retired from the MCP surface. Redundant — agents resolve files via Grep/Read and use validate_symbols/get_neighbors. Internal adapter method stays (drift detection still uses it).

Linked issues

Refs the v0 Productization Notion page (§1 — Allowlist changes during setup).

Plan / Audit / Seal

Plan: trivial; risk:L1.

No schema changes.
No new dependencies.
New code is wizard-only; only the extract_symbols tool surface and the rationale telemetry param are removed (no callers in production code).

Test plan

pytest tests/test_setup_wizard.py -q (11/11)
pytest tests/test_preflight_telemetry.py -q (19/19)
pytest tests/test_phase1_code_locator.py -q (passes — internal extract_symbols still works)
pytest tests/test_hook_command_registration.py -q (passes)
Manual: run bicameral-mcp setup in a scratch dir, accept allowlist, confirm ~/.claude/settings.json has the new entries and project .claude/settings.json is untouched.
Manual: trigger /bicameral-report-bug in Claude Code, verify zero Bash calls and a clickable GitHub-issue URL with dev,bug labels.

🤖 Generated with Claude Code

…-free Telemetry no longer collects the per-skill `rationale` one-liner — it was descriptive freeform text that risked leaking session context into PostHog without measurable analytic value. Removes the `rationale` parameter from `record_skill_event`, the `rationale` property from the `bicameral.skill_begin` tool schema, the `_skill_sessions[...]["rationale"]` plumbing in server.py, and the `rationale=...` examples from every skill's `skill_begin` snippet. Report-bug skill: replaces all Bash blocks (env probing, URL construction via `python3 -c`, `open`/`xdg-open` browser launch) with agent-native behavior — `Read` for `.bicameral/config.yaml`, agent-side URL encoding, and printing the URL for the user to click. Keeps the same redaction rules and the Step 3.5 consent gate. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

coderabbitai · 2026-05-06T01:02:06Z

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: e09300ac-b524-4278-9a2f-6bd4dc4c994b

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch chore/remove-rationale-and-bash-from-report-bug

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

Implements §1 of the v0 productization plan: the wizard offers to write a permissions allowlist into ~/.claude/settings.json so catch-up flows that fire pure-read ledger queries don't spam the user with permission prompts. Bash, Edit, Write, Read/Grep/Glob, and every non-bicameral tool stay prompted — only bicameral MCP tools are pre-approved, and bicameral.reset is deny-listed so it always confirms. Wizard shows the exact diff before writing and requires explicit y/N in interactive runs. Project-level .claude/settings.json is never touched (no commit pollution). Tests cover: user-level write target, no Bash/Edit/Write in the allow list, idempotency (preserves user's own allow entries), declined consent writes nothing, extract_symbols not auto-approved. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Retires the public MCP-tool surface for extract_symbols. The internal adapter method (RealCodeLocatorAdapter.extract_symbols) and the tree-sitter helper (code_locator.indexing.symbol_extractor.extract_symbols) remain — handlers/detect_drift.py still calls them. What's gone is the MCP-callable tool that exposed the same capability to agents: - server.py: header doc count drops 15 → 14, _TOOL_NAMES entry removed, Tool(...) registration removed, dispatch branch removed. - README.md / docs/v0-architecture-current.md: drop extract_symbols from the deterministic-primitives table. - handlers/link_commit.py: tighten the ungrounded-decisions guidance string to point only at validate_symbols. The capability was redundant on the MCP surface — agents resolve files via Grep/Read and use validate_symbols / get_neighbors for index-aware work. Drift detection (the only internal consumer) keeps its direct call into the adapter. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

…icameralAI#200 A4) Closes A4 of BicameralAI#200's audit findings (privacy hardening). The current bug-report skill (post BicameralAI#201) does Read on .bicameral/config.yaml and dumps <contents> verbatim into the issue body. Workspace IDs, tokens, allowlists, and env-specific values leak into GitHub issues — exactly the kind of "presence not value" data the user's "transparency + accuracy + minimum data shared" directive flagged. Changes (skills/bicameral-report-bug/SKILL.md, three edits): 1. Step 2 §config.yaml: default extraction is now top-level keys only (sorted, one per line, no values / nested keys / comments). Sufficient diagnostic signal for "is this bug in the config loader?" while leaking zero values by default. 2. Step 3 body-assembly template: replace the verbatim ```yaml <contents> ``` block with the keys-only shape. Add a "values redacted by default — opt in via Step 3.5 to include verbatim" sentinel line. 3. Step 3.5 transparency preview: add the explicit verbatim toggle as a new option in the AskUserQuestion. When the operator picks "Yes, but include config.yaml verbatim", the body regenerates with the verbatim block and the preview re-displays with the new shape so the operator sees what's actually being shipped before clicking through. Update the Auto-redacted summary block to print the chosen shape. Defense-in-depth: the secret-redaction regex (api_key|token|secret| password|bearer) still runs on verbatim contents. Other findings status: - A1 (python3 portability) — closed by BicameralAI#201 (full bash removal) - A6 (browser-open success) — closed by BicameralAI#201 (URL printed, user clicks) - A7 (telemetry transparency) — partially closed by BicameralAI#201 (rationale field dropped from skill_begin telemetry) Test functionality carve-out justified per doctrine-test-functionality and the precedent across plan-156 PR A Phase 2, plan-156b Phase 1, plan-187 Phase 2, plan-197 Phase 1: skill markdown is LLM-consumed agent-instruction, not pytest-invocable. Plan-grounding lint (BicameralAI#114) self-test exit 0. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

jinhongkuan temporarily deployed to ci-test May 6, 2026 01:02 — with GitHub Actions Inactive

jinhongkuan temporarily deployed to production May 6, 2026 01:02 — with GitHub Actions Inactive

jinhongkuan temporarily deployed to ci-test May 6, 2026 01:02 — with GitHub Actions Inactive

jinhongkuan had a problem deploying to recording-approval May 6, 2026 01:02 — with GitHub Actions Failure

jinhongkuan temporarily deployed to ci-test May 6, 2026 01:02 — with GitHub Actions Inactive

jinhongkuan and others added 2 commits May 5, 2026 18:22

jinhongkuan temporarily deployed to ci-test May 6, 2026 01:23 — with GitHub Actions Inactive

jinhongkuan temporarily deployed to production May 6, 2026 01:23 — with GitHub Actions Inactive

jinhongkuan had a problem deploying to recording-approval May 6, 2026 01:23 — with GitHub Actions Failure

jinhongkuan changed the title ~~chore(telemetry): drop rationale; make report-bug bash-free~~ feat(mcp): MCP UX improvement May 6, 2026

jinhongkuan added the flow:feature Standard feature/fix PR targeting BicameralAI/dev (the default flow) label May 6, 2026

Knapp-Kevin merged commit ddc47ba into dev May 6, 2026
8 of 9 checks passed

Knapp-Kevin deleted the chore/remove-rationale-and-bash-from-report-bug branch May 6, 2026 03:29

Knapp-Kevin mentioned this pull request May 6, 2026

fix(skill): bicameral-report-bug — config.yaml keys-only by default (#200 A4) #204

Merged

3 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(mcp): MCP UX improvement#201

feat(mcp): MCP UX improvement#201
Knapp-Kevin merged 3 commits into
devfrom
chore/remove-rationale-and-bash-from-report-bug

jinhongkuan commented May 6, 2026 •

edited

Loading

Uh oh!

coderabbitai Bot commented May 6, 2026 •

edited

Loading

Review skipped

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

jinhongkuan commented May 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Linked issues

Plan / Audit / Seal

Test plan

Uh oh!

coderabbitai Bot commented May 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review skipped

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

jinhongkuan commented May 6, 2026 •

edited

Loading

coderabbitai Bot commented May 6, 2026 •

edited

Loading