Add ai/recap and ai/security shortcuts

.github/copilot-instructions.md

@@ -189,16 +189,52 @@ Follow the [Stimulus Handbook](https://stimulus.hotwired.dev/handbook/introducti
 - Follow the red-green-refactor cycle: failing test, minimal fix, then refactor
 - Be careful with system/JS tests — avoid patterns that lead to flakiness
 
+## Session recap
+
+When the user says **"recap"**, **"ai recap"**, or runs **`ai/recap`**, review the full conversation and report two parts:
+
+1. **Recap** — what was accomplished this session.
+2. **Unresolved** — dropped threads, unanswered questions, unfinished tasks, and unresolved disagreements (from either side).
+
+This is an agent task — NOT the `/audit` skill (design/accessibility review, only on an explicit `/audit`) and NOT `ai/security` (the security scan). The `ai/recap` script only emits the trigger word; the agent does the work per this section.
+
+**Format the Unresolved part as a bulleted list with a count header.**
+
+If nothing is unresolved:
+```
+- Nothing unresolved
+  - All tasks completed, questions answered, and threads closed
+```
+
+If there are unresolved items:
+```
+- 3 unresolved items below
+  - 1. Item title
+    - Description of what's unresolved
+  - 2. Item title
+    - Description of what's unresolved
+  - 3. Item title
+    - Description of what's unresolved
+```
+
+### After submitting a PR
+
+After creating or submitting a pull request, automatically perform the session recap (Recap + Unresolved) using the format above.
+
 ## Quick Commands
 
 See `ai/` directory for executable scripts:
 
 | Command | What it does |
 |---|---|
+| `ai/recap` | Session recap: accomplishments + unresolved items (see above) |
 | `ai/test [args]` | Run RSpec |
 | `ai/lint` | Rubocop on all files |
 | `ai/lint --fix` | Auto-fix lint issues |
 | `ai/server` | Start dev services (web + vite) |
 | `ai/console` | Rails console |
 | `ai/routes -g pattern` | Search Rails routes |
 | `ai/db-migrate` | Run database migrations |
+| `ai/security` | Security scan: Brakeman + bundler-audit (mirrors CI) |
+
+> **"ai <name>" means the `ai/` script of that name** (e.g. "ai test" → `ai/test`, "ai security" → `ai/security`) — shell scripts in `ai/`, not slash-command skills. If a referenced `ai/<name>` script doesn't exist, ask what's intended rather than substituting a similarly named skill. (`ai/recap` is special — it triggers the agent **Session recap** behavior above, not a real script's output; never confuse it with the `/audit` design skill or the `ai/security` scan.)

CLAUDE.md

@@ -198,16 +198,52 @@ Standard mise activation (`eval "$(mise activate <shell>)"` in your shell rc, pe
 - Follow the red-green-refactor cycle: failing test, minimal fix, then refactor
 - Be careful with system/JS tests — avoid patterns that lead to flakiness
 
+## Session recap
+
+When the user says **"recap"**, **"ai recap"**, or runs **`ai/recap`**, review the full conversation and report two parts:
+
+1. **Recap** — what was accomplished this session.
+2. **Unresolved** — dropped threads, unanswered questions, unfinished tasks, and unresolved disagreements (from either side).
+
+This is an agent task — NOT the `/audit` skill (design/accessibility review, only on an explicit `/audit`) and NOT `ai/security` (the security scan). The `ai/recap` script only emits the trigger word; the agent does the work per this section.
+
+**Format the Unresolved part as a bulleted list with a count header.**
+
+If nothing is unresolved:
+```
+- Nothing unresolved
+  - All tasks completed, questions answered, and threads closed
+```
+
+If there are unresolved items:
+```
+- 3 unresolved items below
+  - 1. Item title
+    - Description of what's unresolved
+  - 2. Item title
+    - Description of what's unresolved
+  - 3. Item title
+    - Description of what's unresolved
+```
+
+### After submitting a PR
+
+After creating or submitting a pull request, automatically perform the session recap (Recap + Unresolved) using the format above.
+
 ## Quick Commands
 
 See `ai/` directory for executable scripts:
 
 | Command | What it does |
 |---|---|
+| `ai/recap` | Session recap: accomplishments + unresolved items (see above) |
 | `ai/test [args]` | Run RSpec |
 | `ai/lint` | Rubocop on all files |
 | `ai/lint --fix` | Auto-fix lint issues |
 | `ai/server` | Start dev services (web + vite) |
 | `ai/console` | Rails console |
 | `ai/routes -g pattern` | Search Rails routes |
 | `ai/db-migrate` | Run database migrations |
+| `ai/security` | Security scan: Brakeman + bundler-audit (mirrors CI) |
+
+> **"ai <name>" means the `ai/` script of that name** (e.g. "ai test" → `ai/test`, "ai security" → `ai/security`) — shell scripts in `ai/`, not slash-command skills. If a referenced `ai/<name>` script doesn't exist, ask what's intended rather than substituting a similarly named skill. (`ai/recap` is special — it triggers the agent **Session recap** behavior above, not a real script's output; never confuse it with the `/audit` design skill or the `ai/security` scan.)

ai/README.md

@@ -4,12 +4,19 @@ Quick-reference scripts for common development tasks. Designed for AI agents and
 
 | Command | What it does |
 |---|---|
+| `ai/recap` | Session recap: accomplishments + unresolved items (agent behavior; see below) |
 | `ai/test [args]` | Run RSpec tests (`ai/test spec/models/user_spec.rb:42`) |
 | `ai/lint` | Rubocop on all files |
 | `ai/lint --fix` | Auto-fix lint issues |
 | `ai/server` | Start all dev services (web + vite) |
 | `ai/console` | Rails console |
 | `ai/routes -g pattern` | Search Rails routes |
 | `ai/db-migrate` | Run database migrations |
+| `ai/security` | Security scan: Brakeman + bundler-audit (mirrors CI) |
 
 All scripts pass through extra arguments, so `ai/test --fail-fast` works as expected.
+
+Only the commands listed above exist. "ai <name>" refers to one of these `ai/` scripts — not a slash-command skill. Two phrases are special:
+
+- **"ai security"** runs `ai/security` (the security scan above).
+- **"ai recap"** (or `ai/recap`, which just prints the trigger word) tells the agent to review the conversation and report two parts: **Recap** (what was accomplished) and **Unresolved** (dropped threads, unanswered questions, unfinished tasks, and disagreements from either side). The agent performs it directly per `CLAUDE.md`. It is never the `/audit` design/accessibility skill.

ai/recap

@@ -0,0 +1,7 @@
+#!/usr/bin/env bash
+# Session recap: triggers the AI to review the conversation and report
+# (1) what was accomplished and (2) unresolved items — dropped threads,
+# unanswered questions, unfinished tasks, and disagreements from either side.
+# This script just prints the trigger word; the agent does the work per CLAUDE.md.
+# Not the /audit design skill, and not ai/security (the security scan).
+echo "recap"

ai/security

@@ -0,0 +1,17 @@
+#!/usr/bin/env bash
+# Security scan: Brakeman (Rails static analysis) + bundler-audit (gem CVEs).
+# Mirrors the CI security scan. Unrelated to the /audit design/accessibility skill.
+# Runs both tools even if the first reports issues, and exits non-zero if either fails.
+set -uo pipefail
+source "$(dirname "$0")/.ruby-env"
+
+status=0
+
+echo "→ Brakeman — Rails security static analysis"
+bin/brakeman --no-pager || status=1
+
+echo
+echo "→ bundler-audit — known CVEs in gems"
+bin/bundler-audit || status=1
+
+exit $status