tech-lead-stack 1.0.3 → 1.1.1

package/.ai/cursor-skills.manifest

@@ -0,0 +1,74 @@
+# Cursor global skills: skill_directory|path relative to tech-lead-stack repo root
+# Lines starting with # are ignored. skill_directory becomes ~/.cursor/skills/<dir>/SKILL.md -> symlink target.
+
+# Core skills (.ai/skills)
+accessibility-auditor|.ai/skills/accessibility-auditor.md
+ask|.ai/skills/ask.md
+capacity-planner|.ai/skills/capacity-planner.md
+changelog-generator|.ai/skills/changelog-generator.md
+clean-code|.ai/skills/clean-code.md
+code-review-checklist|.ai/skills/code-review-checklist.md
+codebase-onboarding-intelligence|.ai/skills/codebase-onboarding-intelligence.md
+competitive-analysis|.ai/skills/competitive-analysis.md
+daily-standup|.ai/skills/daily-standup.md
+design-system-review|.ai/skills/design-system-review.md
+dev-team-local|.ai/skills/dev-team-local.md
+dev-team-orchestrator|.ai/skills/dev-team-orchestrator.md
+dev-team-sub-max|.ai/skills/dev-team-sub-max.md
+dev-team-sub-pro|.ai/skills/dev-team-sub-pro.md
+feature-design-assistant|.ai/skills/feature-design-assistant.md
+feature-orchestrator|.ai/skills/feature-orchestrator.md
+mission-architect|.ai/skills/mission-architect.md
+planning-expert|.ai/skills/planning-expert.md
+planning-expert-quick|.ai/skills/planning-expert-quick.md
+pr-automator|.ai/skills/pr-automator.md
+product-strategist|.ai/skills/product-strategist.md
+qa-handover-generator|.ai/skills/qa-handover-generator.md
+reflexion-loop|.ai/skills/reflexion-loop.md
+reflexion-loop-local|.ai/skills/reflexion-loop-local.md
+reflexion-loop-sub-max|.ai/skills/reflexion-loop-sub-max.md
+reflexion-loop-sub-pro|.ai/skills/reflexion-loop-sub-pro.md
+regression-bug-fix|.ai/skills/regression-bug-fix.md
+security-audit|.ai/skills/security-audit.md
+solutioning-facilitator|.ai/skills/solutioning-facilitator.md
+style-logic-exporter|.ai/skills/style-logic-exporter.md
+technical-debt-auditor|.ai/skills/technical-debt-auditor.md
+ui-spec-generator|.ai/skills/ui-spec-generator.md
+vertical-slice-decomposer|.ai/skills/vertical-slice-decomposer.md
+visual-verifier|.ai/skills/visual-verifier.md
+weekly-leadership-report|.ai/skills/weekly-leadership-report.md
+
+# Antigravity-style workflows (.agents/workflows) — distinct names to avoid clashing with skills
+workflow-accessibility-audit|.agents/workflows/accessibility-audit.md
+workflow-ask|.agents/workflows/ask.md
+workflow-audit-tech-debt|.agents/workflows/audit-tech-debt.md
+workflow-changelog|.agents/workflows/changelog.md
+workflow-clean-code-audit|.agents/workflows/clean-code-audit.md
+workflow-code-review|.agents/workflows/code-review.md
+workflow-competitive-analysis|.agents/workflows/competitive-analysis.md
+workflow-design-requirements-to-architecture|.agents/workflows/design-requirements-to-architecture.md
+workflow-design-system-review|.agents/workflows/design-system-review.md
+workflow-dev-team-sub-max|.agents/workflows/dev-team-sub-max.md
+workflow-dev-team-sub-pro|.agents/workflows/dev-team-sub-pro.md
+workflow-dev-team|.agents/workflows/dev-team.md
+workflow-feature-orchestrator|.agents/workflows/feature-orchestrator.md
+workflow-init|.agents/workflows/init.md
+workflow-mission-architect|.agents/workflows/mission-architect.md
+workflow-onboard-dev|.agents/workflows/onboard-dev.md
+workflow-plan-quick|.agents/workflows/plan-quick.md
+workflow-plan|.agents/workflows/plan.md
+workflow-pr-automator|.agents/workflows/pr-automator.md
+workflow-pr-design-review-init|.agents/workflows/pr-design-review-init.md
+workflow-qa-handover|.agents/workflows/qa-handover.md
+workflow-reflexion-loop-sub-max|.agents/workflows/reflexion-loop-sub-max.md
+workflow-reflexion-loop-sub-pro|.agents/workflows/reflexion-loop-sub-pro.md
+workflow-reflexion-loop|.agents/workflows/reflexion-loop.md
+workflow-regression-bug-fix|.agents/workflows/regression-bug-fix.md
+workflow-security-audit|.agents/workflows/security-audit.md
+workflow-standup-daily-summary|.agents/workflows/standup-daily-summary.md
+workflow-strategy-target-evaluation|.agents/workflows/strategy-target-evaluation.md
+workflow-style-logic-exporter|.agents/workflows/style-logic-exporter.md
+workflow-ui-spec-generator|.agents/workflows/ui-spec-generator.md
+workflow-verify-changes|.agents/workflows/verify-changes.md
+workflow-vertical-slice|.agents/workflows/vertical-slice.md
+workflow-weekly-leadership-report|.agents/workflows/weekly-leadership-report.md

package/.ai/skills/planning-expert.md

@@ -27,7 +27,7 @@ ownership:
   drive: human-ai
   approve: human
 targets: [local, api, subscription]
-minModelClass: small
+minModelClass: large
 consumes: [spec]
 emits: [plan]
 requires: [pr-automator]

package/.ai/skills.graph.json

@@ -1,6 +1,6 @@
 {
   "version": "1.0",
-  "generatedAt": "2026-09-25T12:15:41.470Z",
+  "generatedAt": "2026-10-03T14:02:49.405Z",
   "nodes": [
     {
       "name": "accessibility-auditor",
@@ -482,7 +482,7 @@
         "api",
         "subscription"
       ],
-      "minModelClass": "small",
+      "minModelClass": "large",
       "consumes": [
         "spec"
       ],

package/README.md

@@ -8,7 +8,37 @@ cloned.
 
 Requires Node.js 22.5 or later.
 
-## Connect it to your editor
+## Set it up
+
+```bash
+npx -y tech-lead-stack@1 init
+```
+
+It finds the editors on this computer (Claude Code, Claude Desktop, Cursor,
+Continue, Cline, Gemini), shows what it will change, and asks once. It adds the
+toolbox to each editor, or puts it behind a gateway such as slm-gate if you
+already use one, so no tool is listed twice. It never replaces a connection you
+already have. It also installs:
+
+- Claude Code's `/tls:*` slash commands;
+- Cursor skills and Continue prompts;
+- the workflow files, into the project you run it from (Antigravity and Gemini
+  read them there);
+- RTK, which cuts the tokens your assistant spends reading command output. It is
+  a checked download of a tested version, turned on for Claude Code.
+
+Then it creates your settings file and, if you like, asks for a database address
+and API keys, which are not shown as you type.
+
+`--dry-run` only shows the plan. `--yes` accepts the recommended choices without
+questions, for AI assistants and scripts. `--no-rtk` and `--no-project` skip
+those parts. Restart your editors afterwards. Run `init` again after an upgrade
+to refresh the copied files; any you edited are left alone.
+
+To remove it all: `npx -y tech-lead-stack@1 uninstall` shows what it would
+remove, and `--apply` removes it. Your settings file and RTK are kept.
+
+## Connect it by hand
 
 ### Claude Code
 
@@ -33,11 +63,50 @@ Add this to the client's MCP server configuration:
 
 Then ask your agent to list the available skills.
 
+### Behind a gateway such as slm-gate
+
+If your editor already talks to a gateway, let the gateway start this server
+instead of registering it a second time. For slm-gate, add these two values to
+the `env` block of your existing `slm-gate` entry, then restart your editor:
+
+```json
+"TLS_ADAPTER": "on",
+"DOWNSTREAM_MCP": "{\"command\":\"npx\",\"args\":[\"-y\",\"tech-lead-stack@1\"]}"
+```
+
+The tools then appear under the gateway's name, such as
+`mcp__slm-gate__list_skills`. See
+[Running the stack behind an upstream MCP proxy](https://github.com/bronz3beard/ai.tech-lead-stack/blob/main/docs/mcp-proxy-setup.md#without-a-clone-start-the-stack-with-npx).
+
+## Check your setup
+
+```bash
+npx -y tech-lead-stack@1 doctor
+```
+
+It checks Node.js, your settings file, which features your keys turn on, the
+usage-metrics database, and how each editor on this computer reaches the
+toolbox, then says what to do about anything missing. It only reads; nothing is
+changed. Add `--json` to get the same report for an AI assistant.
+
+## Settings and API keys
+
+Put settings in one file, `~/.tech-lead-stack/.env`, and every editor uses them.
+A value in an editor's MCP `env` block wins over the file, for that editor only.
+The server never reads the `.env` of the project you have open; that file
+belongs to your app.
+
 Tools that call models themselves, such as the reflexion loops, need the
-matching API keys (for example `ANTHROPIC_API_KEY`, `GEMINI_API_KEY`) in the
-server's environment. See
+matching API keys (for example `ANTHROPIC_API_KEY`, `GEMINI_API_KEY`). See
 [Configuration](https://github.com/bronz3beard/ai.tech-lead-stack/blob/main/docs/configuration.md).
 
+## Usage metrics
+
+Add `DATABASE_URL` to `~/.tech-lead-stack/.env` and every skill and tool call is
+recorded. To see the results, run the web app against the same database:
+[Run it on your own machine](https://github.com/bronz3beard/ai.tech-lead-stack/blob/main/docs/web-app.md#run-it-on-your-own-machine).
+Without `DATABASE_URL`, nothing is recorded and everything else works.
+
 ## Verify the package
 
 Every release is built and published by GitHub Actions with npm provenance: