chore: adopt compound-engineering cursor conventions

.agents/docs/README.md

@@ -1,11 +1,11 @@
 # Agent / AI guidance
 
-Project notes for coding agents (Cursor, Claude Code, etc.).
+Durable project context for coding agents (Cursor, Claude Code, etc.). Feature workflow follows Compound Engineering — see [AGENTS.md](../../AGENTS.md).
 
 | Path | Purpose |
 |------|---------|
-| [`../../AGENTS.md`](../../AGENTS.md) | Root agent entrypoint (commands + index) |
-| [`DEVELOPMENT.md`](DEVELOPMENT.md) | TEE trust model, product roles (`BOT__ROLE`), dual-CVM ops |
-| [`../skills/`](../skills/) | Vendored SkillsMP skills (Rust, Docker, Stripe) |
+| [`../../AGENTS.md`](../../AGENTS.md) | Root agent entrypoint (CE loop, commands, index) |
+| [`DEVELOPMENT.md`](DEVELOPMENT.md) | TEE trust model, product roles (`BOT__ROLE`), dual-CVM ops — read when planning TEE/role/compose work |
+| [`../skills/`](../skills/) | Domain skills (Rust, Docker, Stripe); coexist with CE plugin skills |
 
-Product docs for humans stay in [`docs/`](../../docs/). Humans: [`README.md`](../../README.md).
+CE artifacts live under [`docs/brainstorms/`](../../docs/brainstorms/), [`docs/plans/`](../../docs/plans/), [`docs/solutions/`](../../docs/solutions/), and [`todos/`](../../todos/). Product docs for humans stay in [`docs/`](../../docs/). Humans: [`README.md`](../../README.md).

.agents/skills/README.md

@@ -2,6 +2,8 @@
 
 Third-party [SkillsMP](https://skillsmp.com/search) / GitHub skills installed for Cursor via `npx skills add … -a cursor --copy`.
 
+These are **domain** skills (Rust, Docker, Stripe). Feature workflow uses the Compound Engineering marketplace plugin (`/ce-brainstorm`, `/ce-plan`, `/ce-work`, …) — see [AGENTS.md](../../AGENTS.md). Prefer CE for planning/shipping; reach for these skills when writing or reviewing stack-specific code.
+
 **Review before trust** — skills run with full agent permissions. Prefer project docs under `.agents/docs/` and `docs/` for Signal/TEE domain rules.
 
 | Skill | Use for | Source |
@@ -19,4 +21,4 @@ npx skills add rbaumier/skills --skill docker -a cursor --copy -y
 npx skills add midudev/autoskills --skill stripe-best-practices -a cursor --copy -y
 ```
 
-Domain ops (TEE, `BOT__ROLE`, dual CVM): [../docs/DEVELOPMENT.md](../docs/DEVELOPMENT.md). Agent entrypoint: [../../AGENTS.md](../../AGENTS.md).
+Domain ops (TEE, `BOT__ROLE`, dual CVM): [../docs/DEVELOPMENT.md](../docs/DEVELOPMENT.md). Agent entrypoint: [../../AGENTS.md](../../AGENTS.md). Cursor project rules: [../../.cursor/rules/](../../.cursor/rules/).

.compound-engineering/config.local.example.yaml

@@ -0,0 +1,127 @@
+# Compound Engineering -- local config
+# Copy to .compound-engineering/config.local.yaml in your project root.
+# All settings are optional. Invalid values fall through to defaults.
+
+# --- Product pulse ---
+# Settings written by /ce-product-pulse first-run interview. Re-run the skill with
+# argument `setup` or `reconfigure` to edit interactively.
+
+# pulse_product_name: "Spiral"                          # used in report titles (no default)
+# pulse_lookback_default: 24h                           # 1h | 24h | 7d | 30d (default: 24h)
+# pulse_primary_event: "session_started"                # the event that means "user showed up"
+# pulse_value_event: "task_completed"                   # the event that means "user got value"
+# pulse_completion_events: "onboarded,first_purchase"   # comma-separated, 0-3 events
+# pulse_quality_scoring: false                          # true | false (default: false; AI products only)
+# pulse_quality_dimension: "answer accuracy"            # dimension scored 1-5 when pulse_quality_scoring is true
+# pulse_analytics_source: posthog                       # posthog | mixpanel | custom (no default)
+# pulse_tracing_source: sentry                          # sentry | datadog | custom (no default)
+# pulse_payments_source: stripe                         # stripe | custom (no default)
+# pulse_db_enabled: false                               # true | false (default: false; read-only DB if true)
+# pulse_metric_sources: "retention_d7=posthog,nps=delighted"  # strategy-metric -> source overrides; comma-separated 'metric=source' pairs; unlisted metrics fall back to pulse_analytics_source
+# pulse_pending_metrics: "retention_d7,nps"             # comma-separated strategy metrics awaiting instrumentation; render as 'no data'
+# pulse_excluded_metrics: "north_star"                  # comma-separated strategy metrics intentionally not in pulse
+
+# --- Output format ---
+# Per-skill output format default. Selects the exclusive format the artifact
+# is written in: `md` produces a markdown file, `html` produces a single
+# self-contained HTML file. The two are mutually exclusive -- there is no
+# sibling artifact. See DESIGN.md or your agent instructions to influence
+# HTML styling. Precedence: a format request in your prompt for that run wins
+# (e.g. "output:html" or "make it HTML"); a preference you established earlier
+# (in-session, saved to memory, or in your agent instructions) overrides these
+# keys (this config is the persisted fallback); pipeline contexts (e.g., LFG,
+# disable-model-invocation) always force `md`.
+
+# plan_output: html              # md | html (default: md)
+# brainstorm_output: html        # md | html (default: md)
+# ideate_output: md              # md | html (default: html -- ideation docs are human-facing, so HTML is the default; set md to opt out)
+
+# --- Fable elevation (Claude Code only) ---
+# Claude Code ONLY; silently ignored on every other harness (Codex, Cursor, ...).
+# When on, /ce-plan and /ce-brainstorm dispatch their reasoning-heavy step (plan
+# authoring; approach generation) to the Fable model via a subagent, so you get
+# higher-reasoning output even when your session model is cheaper -- no session
+# switch. Per-skill so you can enable one without the other. You can also just
+# say "use fable" in a prompt for a one-off. Invalid/commented/missing -> off.
+
+# plan_use_fable: true           # true | false (default: false)
+# brainstorm_use_fable: true     # true | false (default: false)
+# fable_nudge: false             # true | false (default: true -- one-time tip when elevation is available but off)
+
+# --- PR concept teaching (ce-commit-push-pr) ---
+# When a PR introduces a concept new to the codebase, the generated PR
+# description gains a "New concepts" section teaching it. The section toggle is
+# the single gate: off disables judgment, the section, the /ce-explain offer,
+# and archival. Archival additionally writes the explainer to docs/explainers/
+# and links it from the PR (full-workflow runs only; override per run with
+# archive:on|off). Note: this file is gitignored and per-checkout -- fresh
+# worktrees don't inherit it.
+
+# pr_teaching_section: false     # true | false (default: true)
+# pr_teaching_archive: true      # true | false (default: false; full-workflow runs only)
+
+# --- ce-commit-push-pr babysit handoff ---
+# After a full-workflow PR is opened, ce-commit-push-pr auto-invokes ce-babysit-pr
+# to watch CI + incoming review and drive the PR toward merge-ready. Set false to
+# opt out standing (per-run override: the babysit:off token).
+# auto_babysit: false            # true | false (default: true)
+
+# --- ce-plan scoping confirmation ---
+# By default /ce-plan pauses before research/plan-write to show a scoping
+# summary and wait for you to confirm scope. Set this to skip that wait: ce-plan
+# composes the summary for itself, records any inferred scope under an
+# Assumptions section, announces that it's proceeding, and keeps going. It skips
+# only that confirmation -- genuine blocking questions and the post-plan menu
+# still appear. Override per run with `confirm:auto` (skip) or `confirm:ask`
+# (force the gate on for one run).
+
+# plan_skip_scoping_confirm: true   # true | false (default: false)
+
+# --- Cross-model review preference ---
+# Preferred peer for ce-doc-review and ce-code-review when their conditional
+# cross-model pass activates. Conversation instructions still win, and the
+# skills skip routes that cannot establish the required host/peer distinction.
+
+# cross_model_peer: codex          # codex | claude | grok | cursor | composer
+
+# --- ce-work implementation engine ---
+# Standing, per-checkout preferences for who authors implementation units. Live
+# task/session/project instructions and LFG's provenance-bearing binding may
+# override or narrow these defaults. `off` disables only this standing preference.
+# `prefer` tries candidates in order, then falls back natively with disclosure.
+# `require` asks before native fallback in an interactive ce-work run and blocks
+# without prompting under LFG/headless execution. Harnesses: codex | claude | grok
+# | cursor. Omit model to use that harness's configured default. Composer is a model
+# family through Cursor: { harness: cursor, model: composer }. Do not put CLI flags
+# or commands here. A candidate equivalent to the current host/default is skipped;
+# a different explicit model in the same harness remains eligible.
+
+# work_engine_mode: prefer        # off | prefer | require (default: off)
+# work_engine_preferences:
+#   - harness: cursor
+#     model: composer
+#   - harness: codex
+#     model: "gpt-5.6"
+#   - harness: claude             # configured default model
+
+# --- ce-promote ---
+# Written automatically when you decline the Spiral setup offer in /ce-promote.
+# Suppresses that one-time setup nudge in this project. Remove the key to re-enable.
+
+# ce_promote_spiral_optout: true   # true | (absent) (default: absent -- offer once)
+
+# --- Sweep (ce-sweep) ---
+
+# Feedback sources swept by /ce-sweep. Generic key: other skills may read this list.
+# Each entry: type (slack | github-issues | email), id (short name), target (channel ID /
+# owner/repo / mailbox), ack_action, closeout_action, sensitive (true withholds content
+# from committed state/plan text), approved (standing approval for source-side writes).
+# feedback_sources:
+#   - { type: slack, id: slack-alpha, target: C0XXXXXXX, ack_action: eyes, closeout_action: white_check_mark, sensitive: false, approved: true }
+#   - { type: github-issues, id: gh-issues, target: owner/repo, ack_action: "feedback:ack", closeout_action: "feedback:resolved", sensitive: false, approved: true }
+
+# Sweep-owned settings.
+# sweep_state_path: docs/feedback-sweep/state.yml   # committed (multi-agent) or /tmp path (solo)
+# sweep_ack_cap: 25                                 # max acks per source per run before the circuit breaker
+# sweep_lease_ttl_minutes: 60                       # single-writer lease staleness threshold
+# sweep_shared_branch: false                        # true: push-gated lease for shared-docs-branch topology

.agents/rules/commit-messages.md → .cursor/rules/commit-messages.mdc

@@ -1,3 +1,8 @@
+---
+description: Conventional commit subject line style for this repo
+alwaysApply: true
+---
+
 # Commit message subject lines
 
 When drafting a git commit subject (conventional `type: subject`):

.cursor/rules/compound-engineering.mdc

@@ -0,0 +1,30 @@
+---
+description: Compound Engineering loop, artifact paths, and product constraints
+alwaysApply: true
+---
+
+# Compound Engineering
+
+Prefer the Compound Engineering loop for non-trivial work:
+
+brainstorm → plan → work → simplify → review → compound
+
+Core skills (after installing the marketplace plugin): `/ce-brainstorm`, `/ce-plan`, `/ce-work`, `/ce-simplify-code`, `/ce-code-review`, `/ce-compound`. After non-trivial fixes or features, prefer `/ce-compound` so learnings land in `docs/solutions/`.
+
+## Artifact paths
+
+| Path | Written by |
+|------|------------|
+| `docs/brainstorms/` | `/ce-brainstorm` |
+| `docs/plans/` | `/ce-plan` |
+| `docs/solutions/` | `/ce-compound` |
+| `todos/` | triage / review findings |
+
+Product docs stay as siblings under `docs/` (architecture, voice, translation). Full agent entrypoint: `AGENTS.md`.
+
+## Product constraints
+
+- Required env: `BOT__ROLE=transcription|translation`
+- Do not reintroduce tools, x402, or general chat paths
+- Image digests stay pinned in compose for attestation
+- Before finishing Rust work run `npm run ci` / `pnpm run ci` (never bare `pnpm ci`)

.cursor/settings.json

@@ -0,0 +1,7 @@
+{
+  "plugins": {
+    "compound-engineering": {
+      "enabled": true
+    }
+  }
+}

.gitignore

@@ -50,3 +50,7 @@ credentials.json
 
 # Git worktrees
 .worktrees/
+
+# Compound Engineering machine-local config (keep *.example.yaml tracked)
+.compound-engineering/*.local.yaml
+.compound-engineering/config.local.yaml

AGENTS.md

@@ -6,6 +6,36 @@ Short entrypoint for coding agents. Humans: see [README.md](README.md).
 
 TEE-hosted Signal bots for **voice transcription** and **group translation** (not a general AI chat assistant). Two phone numbers / two bots in a Signal group; Signal is the bus — no Docker network between CVMs. Whisper stays on the transcription stack only.
 
+## Compound Engineering
+
+This repo uses [Compound Engineering](https://every.to/guides/compound-engineering) via the Cursor marketplace plugin (not vendored). Each unit of work should make the next unit easier.
+
+### Install / first run (manual)
+
+In Cursor Agent chat:
+
+```text
+/add-plugin compound-engineering
+/ce-setup
+```
+
+`/ce-setup` should report gitignore coverage for machine-local config, a present `.compound-engineering/config.local.example.yaml`, and no obsolete local md. Optional tools (`gh`, `ast-grep`, etc.) are informational only.
+
+### Core loop
+
+```text
+/ce-brainstorm → /ce-plan → /ce-work → /ce-simplify-code → /ce-code-review → /ce-compound
+```
+
+| Artifact | Path |
+|----------|------|
+| Brainstorms | [`docs/brainstorms/`](docs/brainstorms/) |
+| Plans | [`docs/plans/`](docs/plans/) |
+| Solutions (compounded learnings) | [`docs/solutions/`](docs/solutions/) |
+| Triage / review todos | [`todos/`](todos/) |
+
+After non-trivial fixes or features, prefer `/ce-compound` so the next cycle starts smarter. Domain skills (Rust, Docker, Stripe) under [`.agents/skills/`](.agents/skills/) coexist with CE plugin skills — use CE for feature workflow, domain skills for language/stack craft.
+
 ## Setup / verify
 
 ```bash
@@ -37,13 +67,15 @@ docker compose -f docker/compose.translation.yaml --env-file docker/translation.
 | [`docs/voice-transcription.md`](docs/voice-transcription.md) | Voice transcription product + pairing |
 | [`docs/in-chat-translation.md`](docs/in-chat-translation.md) | In-chat (group) bilingual auto/manual translate |
 | [`docs/language-threads.md`](docs/language-threads.md) | Language Threads (multilingual main + N sidecars) |
-| [`.agents/skills/`](.agents/skills/) | Vendored skills (Rust, Docker, Stripe) |
-| [`.agents/rules/`](.agents/rules/) | Agent rules (e.g. commit message subjects) |
+| [`docs/solutions/`](docs/solutions/) | Compounded learnings from prior work |
+| [`docs/plans/`](docs/plans/) | CE implementation plans |
+| [`.agents/skills/`](.agents/skills/) | Domain skills (Rust, Docker, Stripe) |
+| [`.cursor/rules/`](.cursor/rules/) | Cursor project rules (commits, compound loop) |
 
 ## Rules of thumb
 
 - Required env: `BOT__ROLE=transcription|translation`
 - Do not reintroduce tools, x402, or general chat paths
 - Image digests stay pinned in compose for attestation
-- Commits must pass [commitlint](https://github.com/conventional-changelog/commitlint) (`type: subject`); subject all lowercase, no trailing period, dashes not snake_case — see [`.agents/rules/commit-messages.md`](.agents/rules/commit-messages.md). Run `npm install` or `pnpm install` so husky `commit-msg` / `pre-push` hooks are active
+- Commits must pass [commitlint](https://github.com/conventional-changelog/commitlint) (`type: subject`); subject all lowercase, no trailing period, dashes not snake_case — see [`.cursor/rules/commit-messages.mdc`](.cursor/rules/commit-messages.mdc). Run `npm install` or `pnpm install` so husky `commit-msg` / `pre-push` hooks are active
 - **CI style gates are not optional.** GitHub Actions (`test.yml` + `commitlint.yml`) fails on fmt, Clippy `-D warnings`, llvm-cov ≥90% lines, and conventional commits. Before finishing Rust work run `npm run ci` / `pnpm run ci` (never bare `pnpm ci`). Husky `pre-push` runs that script; `commit-msg` runs commitlint on each commit. Cursor auto-fmts `.rs` edits and re-prompts on stop if fmt/clippy would fail CI.

README.md

@@ -99,7 +99,7 @@ See `.env.example` and the docker `*.env.example` files.
 
 See [.agents/docs/DEVELOPMENT.md](.agents/docs/DEVELOPMENT.md) for the TEE trust model, why Signal CLI must run in the TEE, and attestation (`!verify`).
 
-For coding agents, see [AGENTS.md](AGENTS.md).
+For coding agents, see [AGENTS.md](AGENTS.md) (includes Compound Engineering install: `/add-plugin compound-engineering`, then `/ce-setup`).
 
 ## License
 

docs/DESIGN.md

@@ -6,7 +6,7 @@ Current product direction:
 
 - [two-cvm-architecture.md](two-cvm-architecture.md)
 - [language-threads.md](language-threads.md)
-- [../.agents/DEVELOPMENT.md](../.agents/DEVELOPMENT.md)
+- [../.agents/docs/DEVELOPMENT.md](../.agents/docs/DEVELOPMENT.md)
 - [../README.md](../README.md)
 - [issue #10](https://github.com/BreadchainCoop/sigstack-bot/issues/10)
 

docs/brainstorms/README.md

@@ -0,0 +1,5 @@
+# Brainstorms
+
+Requirements-oriented brainstorm docs written by `/ce-brainstorm` (Compound Engineering).
+
+Hand off approved brainstorms into `/ce-plan`. Product specs live alongside this folder under [`docs/`](../).

docs/plans/README.md

@@ -0,0 +1,5 @@
+# Plans
+
+Implementation-ready plans written by `/ce-plan` (Compound Engineering).
+
+Product docs stay as siblings under [`docs/`](../). Do not put durable product specs here — use the named product markdown files instead.

docs/solutions/README.md

@@ -0,0 +1,5 @@
+# Solutions
+
+Institutional learnings written by `/ce-compound` (Compound Engineering).
+
+Each solved problem becomes searchable markdown so the next `/ce-brainstorm` / `/ce-plan` cycle starts smarter. Prefer compounding after non-trivial fixes and features.

todos/README.md

@@ -0,0 +1,5 @@
+# Todos
+
+Triage and review findings from Compound Engineering workflows (`/triage`, `/ce-code-review` resolution).
+
+Files are typically named with priority and status (for example `001-ready-p1-…`). Work ready items with CE resolve skills; do not treat this folder as a product backlog.