@techierathore/techieflow 1.0.2 → 1.0.3

package/docs/TechieFlow-Installation.md

@@ -17,7 +17,7 @@ You need four things on the machine.
 |---|---|---|
 | Node.js 20 or newer | Runs the installer. Nothing is added to your project. | `node --version` |
 | bash | The framework's guard hooks and helper scripts run under bash. macOS and Linux have it. On Windows, work inside WSL or Git Bash. | `bash --version` |
-| Python 3.10 or newer | Powers the HTML renderer, the telemetry writer and the guard hooks. An older Python 3 works but skips the Codex files. | `python3 --version` |
+| Python 3 | Powers the HTML renderer, the telemetry writer and the guard hooks. | `python3 --version` |
 | Claude Code or OpenCode | The harness that runs the agents. Either one. Both work from the same install. | `claude --version` or `opencode --version` |
 
 For a .NET project you also need the .NET SDK. Run `dotnet --version` to check.
@@ -42,8 +42,6 @@ The npm package itself holds the framework folders, this document and a `scripts
 | `.opencode/opencode.jsonc` | The framework's OpenCode configuration. OpenCode reads the personas and tasks straight from `.tfcore/` through it. | Refreshed. |
 | `.opencode/command/generate-html.md` | The short `/generate-html` command for OpenCode. | Added on update only. |
 | `opencode.jsonc` | The root OpenCode configuration, for your own additions. | Refreshed only when it holds nothing of yours, with the old file kept as `opencode.jsonc.bak`. |
-| `.codex/` and `.agents/skills/` | The Codex adapter, generated from the same personas and tasks. | Refreshed. `config.toml` is yours and is kept. |
-| `WORKFLOW.html` | The human workflow guide. Open it in a browser. | Refreshed. |
 | `.tf-scaffold-note.txt` | A note with your next command. Delete it when read. | Left alone. |
 | `docs/metrics/` | Five empty telemetry files and a README. This is your project's history: commit it. | Left alone. |
 | `.gitignore` | Lines that keep the copies above out of your commits. Appended, never rewritten. | Appended. |
@@ -242,7 +240,6 @@ The installer and the updater never write to:
 - `.tfcore/core-config.yaml` and `.tfcore/routing.yaml` after the first install
 - `.claude/settings.local.json`
 - `.claude/commands/trblazeui.md`, `.claude/commands/techierag.md` and the same names under `.opencode/command/`. These come from the NuGet packages.
-- `.codex/config.toml` after the first install
 - `opencode.jsonc` at the root when it holds a key of your own
 
 Existing lines in `.gitignore` and `.gitattributes` are never rewritten. The managed lines are appended once and recognised on every later run.
@@ -316,7 +313,6 @@ It installs the framework if it is not there yet, then removes the leftover pack
 | The framework is under `node_modules/@techierathore/techieflow/` and nowhere else | You ran `npm install`. See section 9. Run `npx @techierathore/techieflow@latest install`. |
 | `bash was not found` | On Windows, run the command inside WSL or Git Bash. |
 | `python3 was not found` | Install Python 3 and run the command again. On macOS: `brew install python3`. On Ubuntu or WSL: `sudo apt-get install -y python3`. |
-| `Codex bindings could not be generated` | Your Python is older than 3.10. Everything except the Codex files is installed. Upgrade Python and run `update` to add them. |
 | `Refusing to install into the framework itself` | You ran the command inside a clone of this repository. Pass `--target=<your project>`. |
 | `does not look installed` on update | The folder has no `.tfcore/`. Run `install` first. |
 | Claude Code does not show the `/TechieFlow:agents:analyst` command | Restart Claude Code in the project folder. Check that `.claude/commands/TechieFlow/agents/analyst.md` exists. |

package/opencode.jsonc

@@ -253,7 +253,7 @@
           "sudo *": "ask"
         }
       },
-      "description": "Madhav — the single super-agent. Orchestrates the full pipeline (*run-workflow, *phase) with parallel subagents, runs the unified build (*build-phase — calls /trblazeui + /techierag as sub-agents), fixes bugs from screenshots (*fix-issues), analyzes + logs human-found bugs without fixing (*triage-issues), records a single missed requirement in seconds (*log-miss), reports development telemetry (*metrics), and handles doc/status utilities (*render-workflow-docs, *generate-html, *handoff-phase, *refresh-status, *devguide) plus any one-off task."
+      "description": "Madhav — the TechieFlow master: build (*build-phase), fix (*fix-issues), triage (*triage-issues), log a miss (*log-miss), amend documents (*amend-docs), DevGuide and ProductGuide, handoff, refresh-status, metrics, HTML rendering, YOLO. The analyst owns day-1; the verifier owns *verify.",
     },
     "flow-architect": {
       "prompt": "{file:./.tfcore/agents/architect.md}",
@@ -839,19 +839,23 @@
     },
     "techieflow:tasks:fix-issues": {
       "template": "{file:./.tfcore/tasks/fix-issues.md}",
-      "description": "Bug-fix front door: given a folder of screenshots (+ optional description), reproduce with Playwright, triage (layout/data/logic/RAG), fan fixes to the right builder, re-smoke (data+visual) + re-verify, update docs. Driven by /flow-master."
+      "description": "Fixes the bugs in a folder of screenshots: triages and logs them, builds the fix through the builders in FIX mode, smokes, re-verifies the touched rows, closes the misses with the fix cost (tf-fix-close.sh). Driven by /flow-master."
     },
     "techieflow:tasks:log-miss": {
       "template": "{file:./.tfcore/tasks/log-miss.md}",
-      "description": "The 20-second front door for 'you missed this': turns one sentence about something an agent got wrong into a misses.jsonl record (what was missed, which phase/agent/model let it through, who found it) plus the checklist line — demotion with a dated miss Remark, or a new Planned row when nothing owns it. Never boots the app, never reproduces, never edits code. Driven by /flow-master."
+      "description": "The twenty-second record: one sentence from the owner becomes one miss record and one checklist line through tf-log-miss.sh. Never boots, never fixes. Driven by /flow-master."
     },
     "techieflow:tasks:triage-issues": {
       "template": "{file:./.tfcore/tasks/triage-issues.md}",
-      "description": "ANALYZE-ONLY bug front door for human-found (UAT/production) bugs: reproduce with Playwright, triage to the owning REQ, log in the checklist (demote to Needs re-verify / new Planned rows), optional scoped regression re-verify, update PROJECT-STATUS with the *fix-issues pointer. NEVER edits code. Driven by /flow-master."
+      "description": "Analyse-only front door for bugs a person found: reproduces each on its screen, logs it in the checklist and the telemetry through tf-triage.sh (demote, new, note, close), never edits code. Driven by /flow-master."
+    },
+    "techieflow:tasks:triage-and-fix": {
+      "template": "{file:./.tfcore/tasks/triage-and-fix.md}",
+      "description": "The owner's whole bug sequence in YOLO: compare every screen to its mockup, triage the reported bugs, log every root cause with its cost, fix, log every fix with its cost, refresh the metrics, one summary per step. Driven by /flow-master."
     },
     "techieflow:tasks:verify-phase": {
       "template": "{file:./.tfcore/tasks/verify-phase.md}",
-      "description": "Autonomous verification of a scope (ui | functional | all | REQ list | legacy phase-N) — filters the one Checklist by REQ prefix; applies the data-render + visual-truth gates; verdicts land in the Checklist Requirements Status table."
+      "description": "Verifies every checklist row in a scope (ui | functional | all | REQ list) against the running app: boots it, runs the seven checks in order through the tf-verify-* scripts, writes the verdicts into the checklist and the telemetry. YOLO by default. Driven by /verifier."
     },
     "techieflow:tasks:handoff-phase": {
       "template": "{file:./.tfcore/tasks/handoff-phase.md}",
@@ -873,14 +877,14 @@
       "template": "{file:./.tfcore/tasks/generate-html.md}",
       "description": "Render any markdown file(s) or a non-recursive directory to self-contained HTML using the shared shell."
     },
-    "techieflow:tasks:author-brd": {
-      "template": "{file:./.tfcore/tasks/author-brd.md}",
-      "description": "Interactively extend docs/{AppName}-BRD.md with confirmed BRD-N requirements (per-item elicitation)."
-    },
     "techieflow:tasks:amend-docs": {
       "template": "{file:./.tfcore/tasks/amend-docs.md}",
       "description": "Fold an evolving concept / changed requirements into the EXISTING day-1 docs IN PLACE: surgically amends BRD + Architecture (append-only IDs), ripples to PROJECT-STATUS / BRD §4 / checklists, re-renders HTML. Incremental alternative to re-running day1-*."
     },
+    "techieflow:tasks:deploy-checklist": {
+      "template": "{file:./.tfcore/tasks/deploy-checklist.md}",
+      "description": "Deployment Checklist: the steps to put the application on its host, one document per hosting target, written after UAT from the owner's pipeline guidance document and the Stack answers Q9 and Q10 (docs/TechieFlow-Document-Schemas.md §3.10). Never deploys. Driven by /flow-master."
+    },
     "techieflow:tasks:productguide": {
       "template": "{file:./.tfcore/tasks/productguide.md}",
       "description": "End-user Product Guide: screenshot-illustrated, task-oriented how-to manual for external users (what each screen is for + how to do things) — the user-facing sibling of the DevGuide, reusing its captured screenshots. Always MD + HTML. On-demand. Driven by /flow-master."
@@ -889,10 +893,6 @@
       "template": "{file:./.tfcore/tasks/devguide.md}",
       "description": "Generate/refresh the screen-by-screen Developer Guide: traces every screen/control from Razor page → service → data-access → stored proc/query, per user role, documenting the code AS BUILT. Single doc for small apps, split per role for large ones; --update refreshes only changed screens."
     },
-    "techieflow:tasks:document-project": {
-      "template": "{file:./.tfcore/tasks/document-project.md}",
-      "description": "Generate comprehensive documentation for existing projects optimized for AI development agents."
-    },
     "techieflow:tasks:create-doc": {
       "template": "{file:./.tfcore/tasks/create-doc.md}",
       "description": "Create a document from a YAML-driven TechieFlow template (interactive, section by section)."
@@ -904,26 +904,6 @@
     "techieflow:tasks:facilitate-brainstorming-session": {
       "template": "{file:./.tfcore/tasks/facilitate-brainstorming-session.md}",
       "description": "Run a structured brainstorming session; results land in docs/brainstorming-session-results.md."
-    },
-    "techieflow:tasks:advanced-elicitation": {
-      "template": "{file:./.tfcore/tasks/advanced-elicitation.md}",
-      "description": "Reflective/brainstorming actions for deeper exploration and iterative refinement of drafted content."
-    },
-    "techieflow:tasks:execute-checklist": {
-      "template": "{file:./.tfcore/tasks/execute-checklist.md}",
-      "description": "Validate a document or work product against one of the TechieFlow checklists."
-    },
-    "techieflow:tasks:index-docs": {
-      "template": "{file:./.tfcore/tasks/index-docs.md}",
-      "description": "Maintain docs/index.md by scanning all documentation files and indexing them with descriptions."
-    },
-    "techieflow:tasks:kb-mode-interaction": {
-      "template": "{file:./.tfcore/tasks/kb-mode-interaction.md}",
-      "description": "User-friendly interface to the TechieFlow knowledge base."
-    },
-    "techieflow:tasks:shard-doc": {
-      "template": "{file:./.tfcore/tasks/shard-doc.md}",
-      "description": "Split a large document into smaller documents by level-2 sections, preserving content integrity."
     }
   }
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@techierathore/techieflow",
-  "version": "1.0.2",
+  "version": "1.0.3",
   "description": "TechieFlow: a spec-driven development framework for AI coding agents, for Claude Code and OpenCode. One command installs it into any project.",
   "license": "Apache-2.0",
   "repository": {
@@ -33,11 +33,7 @@
     ".opencode/command",
     "!.opencode/command/trblazeui.md",
     "!.opencode/command/techierag.md",
-    ".codex/config.toml",
-    ".codex/hooks.json",
-    ".codex/rules",
     "opencode.jsonc",
-    "WORKFLOW.html",
     "scripts/install.mjs",
     "scripts/npm-postinstall.mjs",
     "scripts/npm-cleanup.mjs",

package/scripts/install.mjs

@@ -20,7 +20,7 @@
 // scripts/test-install.mjs proves that by running both routes and comparing the results.
 //
 // Needs: node, bash and python3. The framework's own helper scripts (telemetry setup,
-// build-output ignore audit, Codex bindings) are run exactly as the shell scripts run them.
+// build-output ignore audit) are run exactly as the shell scripts run them.
 
 import { spawnSync } from "node:child_process";
 import {
@@ -201,7 +201,7 @@ function checkTools() {
   }
   pythonCommand = findPython();
   if (!pythonCommand) {
-    throw new Error("python3 was not found. It powers the HTML renderer, the telemetry writer, the Codex bindings and every guard hook.\n"
+    throw new Error("python3 was not found. It powers the HTML renderer, the telemetry writer and every guard hook.\n"
       + "  Install it and run this command again:\n"
       + "    macOS:          brew install python3\n"
       + "    Ubuntu / WSL:   sudo apt-get install -y python3\n"
@@ -237,8 +237,8 @@ function ensureSafeTarget() {
 // Same lines, same patterns and same header text as the shell scripts. Append-only: an
 // entry already present in any anchored or slash variant is respected, nothing is rewritten.
 const frameworkIgnore = {
-  lines: [".tfcore/", ".claude/", ".opencode/", ".codex/", ".agents/skills/", "/CLAUDE.md", "/WORKFLOW.html", "/opencode.jsonc", "/.tf-scaffold-note.txt"],
-  patterns: [/^\/?\.tfcore\/?$/, /^\/?\.claude\/?$/, /^\/?\.opencode\/?$/, /^\/?\.codex\/?$/, /^\/?\.agents\/skills\/?$/, /^\/?CLAUDE\.md$/, /^\/?WORKFLOW\.html$/, /^\/?opencode\.jsonc$/, /^\/?\.tf-scaffold-note\.txt$/],
+  lines: [".tfcore/", ".claude/", ".opencode/", "/CLAUDE.md", "/opencode.jsonc", "/.tf-scaffold-note.txt"],
+  patterns: [/^\/?\.tfcore\/?$/, /^\/?\.claude\/?$/, /^\/?\.opencode\/?$/, /^\/?CLAUDE\.md$/, /^\/?opencode\.jsonc$/, /^\/?\.tf-scaffold-note\.txt$/],
   header: ["# TechieFlow framework — deployed copies, never commit (managed by scaffold/update-framework.sh)"],
   label: "framework entries",
 };
@@ -353,6 +353,22 @@ const canonicalSettings = String.raw`{
           {
             "type": "command",
             "command": "bash \"$CLAUDE_PROJECT_DIR/.tfcore/hooks/guard-artifacts.sh\""
+          },
+          {
+            "type": "command",
+            "command": "bash \"$CLAUDE_PROJECT_DIR/.tfcore/hooks/guard-status.sh\""
+          },
+          {
+            "type": "command",
+            "command": "bash \"$CLAUDE_PROJECT_DIR/.tfcore/hooks/guard-metrics.sh\""
+          },
+          {
+            "type": "command",
+            "command": "bash \"$CLAUDE_PROJECT_DIR/.tfcore/hooks/guard-db.sh\""
+          },
+          {
+            "type": "command",
+            "command": "bash \"$CLAUDE_PROJECT_DIR/.tfcore/hooks/guard-build.sh\""
           }
         ]
       },
@@ -363,6 +379,10 @@ const canonicalSettings = String.raw`{
             "type": "command",
             "command": "bash \"$CLAUDE_PROJECT_DIR/.tfcore/hooks/guard-status.sh\""
           },
+          {
+            "type": "command",
+            "command": "bash \"$CLAUDE_PROJECT_DIR/.tfcore/hooks/guard-metrics.sh\""
+          },
           {
             "type": "command",
             "command": "bash \"$CLAUDE_PROJECT_DIR/.tfcore/hooks/guard-verify.sh\""
@@ -426,7 +446,6 @@ Added (only missing files filled — re-runs are safe):
   .tfcore/                  ← TechieFlow v4 customized (agents, tasks, templates)
   .claude/commands/            ← Claude Code slash commands (TechieFlow agents)
   .claude/settings.json        ← yolo-except-git-writes permissions
-  WORKFLOW.html                ← the human workflow guide (open in a browser; §17 = macOS / Windows / Linux)
   opencode.jsonc               ← OpenCode config (loads agents/tasks from .tfcore/ via {file:...} refs)
   .gitignore                   ← framework entries appended (deployed copies stay uncommitted)
 
@@ -468,7 +487,6 @@ Folders/files created (only missing files filled — re-runs are safe):
   .tfcore/                  ← TechieFlow v4 customized (agents, tasks, templates)
   .claude/commands/            ← Claude Code slash commands (TechieFlow agents)
   .claude/settings.json        ← yolo-except-git-writes permissions
-  WORKFLOW.html                ← the human workflow guide (open in a browser; §17 = macOS / Windows / Linux)
   opencode.jsonc               ← OpenCode config (loads agents/tasks from .tfcore/ via {file:...} refs)
   .gitignore                   ← framework entries appended (deployed copies stay uncommitted)
   tests/playwright/  tests/unit/  src/
@@ -515,15 +533,20 @@ function deployOpenCodeBridge() {
   writeText(join(target, ".opencode", "opencode.jsonc"), openCodeConfigCopy());
 }
 
-function deployCodexAdapter() {
-  say("  .codex/ + .agents/skills/ — Codex adapter");
-  if (dryRun) { say("  WOULD preserve/create .codex/config.toml; refresh hooks/rules; regenerate agents/skills"); return; }
-  for (const folder of [".codex/agents", ".codex/rules", ".agents/skills"]) mkdirSync(join(target, folder), { recursive: true });
-  copyFile(join(sourceRoot, ".codex", "config.toml"), join(target, ".codex", "config.toml"), { onlyIfMissing: true });
-  copyFile(join(sourceRoot, ".codex", "hooks.json"), join(target, ".codex", "hooks.json"));
-  copyFile(join(sourceRoot, ".codex", "rules", "techieflow.rules"), join(target, ".codex", "rules", "techieflow.rules"));
-  const result = runPython([join(target, ".tfcore", "utils", "tf-codex-bind.py"), target]);
-  if (result.status !== 0) say("  ⚠ Codex bindings could not be generated (python3 required)");
+// The Codex adapter was removed on 2026-09-07 (D-14, FR-42) and WORKFLOW.html was dropped the
+// same day: it was a second full description of the process, last revised before the reset, and
+// it still taught commands that no longer exist. Both are taken out of a project that has them.
+// Nothing here is project content: .codex/ held a config file plus generated bindings,
+// .agents/skills/ was generated in full from .tfcore/tasks/, and WORKFLOW.html was a copy.
+function removeRetiredFiles() {
+  for (const path of [".codex", ".agents/skills", "WORKFLOW.html"]) {
+    const full = join(target, path);
+    if (!existsSync(full)) continue;
+    if (dryRun) { say(`  ${would}remove ${path} — no longer part of the framework`); continue; }
+    rmSync(full, { recursive: true, force: true });
+    say(`  removed ${path} — no longer part of the framework`);
+  }
+  if (!dryRun) { try { rmdirSync(join(target, ".agents")); say("  removed the emptied .agents/"); } catch {} }
 }
 
 function deployHousekeeping() {
@@ -619,13 +642,11 @@ async function install() {
   say("  syncing agent files from .tfcore/agents/ → .claude/commands/TechieFlow/agents/");
   copyTree(dryRun ? join(sourceRoot, ".tfcore", "agents") : join(target, ".tfcore", "agents"), join(target, ".claude", "commands", "TechieFlow", "agents"));
   // 4. reference files at the root, only if missing
-  copyFile(join(sourceRoot, "WORKFLOW.html"), join(target, "WORKFLOW.html"), { onlyIfMissing: true });
   copyFile(join(sourceRoot, "opencode.jsonc"), join(target, "opencode.jsonc"), { onlyIfMissing: true });
   // 4b. OpenCode bridge, always refreshed
   deployOpenCodeBridge();
-  // 4c. Codex adapter
-  deployCodexAdapter();
-  if (!dryRun) say("  Codex adapter installed — trust this repository and review /hooks before relying on guards");
+  // 4c. files earlier versions deployed and the framework no longer ships
+  removeRetiredFiles();
   // 5. Claude Code permissions, only if missing
   if (writeText(join(target, ".claude", "settings.json"), `${canonicalSettings}\n`, { onlyIfMissing: true })) { if (!dryRun) say("  created .claude/settings.json"); }
   else say("  .claude/settings.json already exists — preserved");
@@ -650,13 +671,15 @@ async function install() {
   say("");
   say(`Next: cd "${target}"`);
   if (greenfield) say("      dotnet new sln + blazor + add TrBlazeUI/TechieRag NuGets + dotnet build");
-  say("      open WORKFLOW.html in a browser");
   say(`      start Claude Code or OpenCode and run the ${greenfield ? "greenfield" : "brownfield"} day-1 command from .tf-scaffold-note.txt`);
 }
 
 // ---------------------------------------------------------------- update (update-framework.sh)
 
-const frameworkSubdirs = ["agents", "tasks", "telemetry", "templates", "checklists", "data", "utils", "hooks", "workflows", "agent-teams"];
+// Keep this in step with FRAMEWORK_SUBDIRS in update-framework.sh. `standards` arrived with the
+// document schemas and was missing here, so a project migrated from the old layout came out
+// without its coding standards (found by scripts/test-install.mjs, 2026-09-07).
+const frameworkSubdirs = ["agents", "tasks", "telemetry", "templates", "checklists", "data", "utils", "hooks", "standards", "workflows", "agent-teams"];
 const frameworkTopFiles = ["enhanced-ide-development-workflow.md", "user-guide.md", "working-in-the-brownfield.md", "install-manifest.yaml", "TOKEN-GUIDE.md"];
 const legacyConfigPattern = /\.bmad-core|bmad-(master|orchestrator|analyst|architect|verifier)|BMad:/;
 
@@ -866,16 +889,10 @@ function update() {
   deployOpenCodeBridge();
   refreshRootOpenCodeConfig();
 
-  // 4b. Codex adapter
-  deployCodexAdapter();
-  if (!dryRun) say("  Codex hooks changed or installed — trust this repository and review /hooks");
+  // 4b. files earlier versions deployed and the framework no longer ships
+  removeRetiredFiles();
   for (const file of libraryPersonas) if (existsSync(join(target, ".opencode", "command", file))) say(`  .opencode/command/${file} — preserved (NuGet-deployed)`);
 
-  // 4. WORKFLOW.html, always refreshed
-  if (existsSync(join(sourceRoot, "WORKFLOW.html"))) {
-    copyFile(join(sourceRoot, "WORKFLOW.html"), join(target, "WORKFLOW.html"));
-    say("  WORKFLOW.html");
-  }
   // 5. NuGet persona shims: gap-fill only, never an overwrite
   shimLegacyPersonas({ gapOnly: true });
   // 6, 7. messages about stale references in files the owner owns
@@ -924,29 +941,18 @@ function uninstall() {
   plan(".opencode/opencode.jsonc");
   plan(".opencode/opencode.json", "(generated by routing)");
   for (const file of shippedFiles(".opencode/command", ".md")) if (!libraryPersonas.has(file)) plan(`.opencode/command/${file}`);
-  plan(".codex/hooks.json");
-  plan(".codex/rules/techieflow.rules");
-  if (existsSync(join(target, ".codex", "config.toml"))) {
-    if (sameAsShipped(".codex/config.toml", ".codex/config.toml")) plan(".codex/config.toml");
-    else keep(".codex/config.toml", "you changed it");
-  }
-  const codexAgents = join(target, ".codex", "agents");
-  if (existsSync(codexAgents)) for (const f of readdirSync(codexAgents)) {
-    if (f.endsWith(".toml") && /description = "TechieFlow [a-z-]+ (specialist|role)\."/.test(read(join(codexAgents, f)))) plan(`.codex/agents/${f}`);
-    else keep(`.codex/agents/${f}`, "not written by the framework");
-  }
-  const skills = join(target, ".agents", "skills");
-  if (existsSync(skills)) for (const f of readdirSync(skills)) if (f.startsWith("techieflow-")) plan(`.agents/skills/${f}`);
+  // Retired on 2026-09-07 but still present in a project set up by an older version.
+  for (const path of [".codex", ".agents/skills"]) if (existsSync(join(target, path))) plan(path, "(retired Codex adapter)");
   for (const lib of [".trblazeui", ".techierag"]) {
     if (!existsSync(join(target, lib))) continue;
     const rest = readdirSync(join(target, lib)).filter((f) => f !== ".codex-agent-package-owned" && f !== ".gitignore");
     plan(`${lib}/.codex-agent-package-owned`);
     if (rest.length === 0) plan(`${lib}/.gitignore`);
   }
-  for (const [path, shipped] of [["WORKFLOW.html", "WORKFLOW.html"], ["opencode.jsonc", "opencode.jsonc"]]) {
-    if (!existsSync(join(target, path))) continue;
-    if (sameAsShipped(path, shipped)) plan(path);
-    else keep(path, "you changed it");
+  if (existsSync(join(target, "WORKFLOW.html"))) plan("WORKFLOW.html", "(retired human workflow guide)");
+  if (existsSync(join(target, "opencode.jsonc"))) {
+    if (sameAsShipped("opencode.jsonc", "opencode.jsonc")) plan("opencode.jsonc");
+    else keep("opencode.jsonc", "you changed it");
   }
   plan(".tf-scaffold-note.txt");
   const hook = join(target, ".git", "hooks", "pre-commit");

package/.claude/commands/TechieFlow/tasks/advanced-elicitation.md

@@ -1,117 +0,0 @@
-# Advanced Elicitation Task
-
-## Purpose
-
-- Provide optional reflective and brainstorming actions to enhance content quality
-- Enable deeper exploration of ideas through structured elicitation techniques
-- Support iterative refinement through multiple analytical perspectives
-- Usable during template-driven document creation or any chat conversation
-
-## Usage Scenarios
-
-### Scenario 1: Template Document Creation
-
-After outputting a section during document creation:
-
-1. **Section Review**: Ask user to review the drafted section
-2. **Offer Elicitation**: Present 9 carefully selected elicitation methods
-3. **Simple Selection**: User types a number (0-8) to engage method, or 9 to proceed
-4. **Execute & Loop**: Apply selected method, then re-offer choices until user proceeds
-
-### Scenario 2: General Chat Elicitation
-
-User can request advanced elicitation on any agent output:
-
-- User says "do advanced elicitation" or similar
-- Agent selects 9 relevant methods for the context
-- Same simple 0-9 selection process
-
-## Task Instructions
-
-### 1. Intelligent Method Selection
-
-**Context Analysis**: Before presenting options, analyze:
-
-- **Content Type**: Technical specs, user stories, architecture, requirements, etc.
-- **Complexity Level**: Simple, moderate, or complex content
-- **Stakeholder Needs**: Who will use this information
-- **Risk Level**: High-impact decisions vs routine items
-- **Creative Potential**: Opportunities for innovation or alternatives
-
-**Method Selection Strategy**:
-
-1. **Always Include Core Methods** (choose 3-4):
-   - Expand or Contract for Audience
-   - Critique and Refine
-   - Identify Potential Risks
-   - Assess Alignment with Goals
-
-2. **Context-Specific Methods** (choose 4-5):
-   - **Technical Content**: Tree of Thoughts, ReWOO, Meta-Prompting
-   - **User-Facing Content**: Agile Team Perspective, Stakeholder Roundtable
-   - **Creative Content**: Innovation Tournament, Escape Room Challenge
-   - **Strategic Content**: Red Team vs Blue Team, Hindsight Reflection
-
-3. **Always Include**: "Proceed / No Further Actions" as option 9
-
-### 2. Section Context and Review
-
-When invoked after outputting a section:
-
-1. **Provide Context Summary**: Give a brief 1-2 sentence summary of what the user should look for in the section just presented
-
-2. **Explain Visual Elements**: If the section contains diagrams, explain them briefly before offering elicitation options
-
-3. **Clarify Scope Options**: If the section contains multiple distinct items, inform the user they can apply elicitation actions to:
-   - The entire section as a whole
-   - Individual items within the section (specify which item when selecting an action)
-
-### 3. Present Elicitation Options
-
-**Review Request Process:**
-
-- Ask the user to review the drafted section
-- In the SAME message, inform them they can suggest direct changes OR select an elicitation method
-- Present 9 intelligently selected methods (0-8) plus "Proceed" (9)
-- Keep descriptions short - just the method name
-- Await simple numeric selection
-
-**Action List Presentation Format:**
-
-```text
-**Advanced Elicitation Options**
-Choose a number (0-8) or 9 to proceed:
-
-0. [Method Name]
-1. [Method Name]
-2. [Method Name]
-3. [Method Name]
-4. [Method Name]
-5. [Method Name]
-6. [Method Name]
-7. [Method Name]
-8. [Method Name]
-9. Proceed / No Further Actions
-```
-
-**Response Handling:**
-
-- **Numbers 0-8**: Execute the selected method, then re-offer the choice
-- **Number 9**: Proceed to next section or continue conversation
-- **Direct Feedback**: Apply user's suggested changes and continue
-
-### 4. Method Execution Framework
-
-**Execution Process:**
-
-1. **Retrieve Method**: Access the specific elicitation method from the elicitation-methods data file
-2. **Apply Context**: Execute the method from your current role's perspective
-3. **Provide Results**: Deliver insights, critiques, or alternatives relevant to the content
-4. **Re-offer Choice**: Present the same 9 options again until user selects 9 or gives direct feedback
-
-**Execution Guidelines:**
-
-- **Be Concise**: Focus on actionable insights, not lengthy explanations
-- **Stay Relevant**: Tie all elicitation back to the specific content being analyzed
-- **Identify Personas**: For multi-persona methods, clearly identify which viewpoint is speaking
-- **Maintain Flow**: Keep the process moving efficiently

package/.claude/commands/TechieFlow/tasks/author-brd.md

@@ -1,79 +0,0 @@
-# author-brd
-
-Author or extend a Business Requirements Document (BRD) with stable BRD-N requirement IDs, using interactive elicitation. The analyst NEVER assumes a requirement — each candidate is confirmed by the user before it lands in the BRD.
-
-> **When to use:** this is the INTERACTIVE alternative for extending an existing BRD requirement-by-requirement. For creating a project's BRD, the default path is `*day1-brownfield` / `*day1-greenfield` (bulk draft, no per-item confirmation).
-
-## Purpose
-
-Produce a BRD where every business/functional requirement has a unique, stable, append-only ID (`BRD-1`, `BRD-2`, …) that downstream agents (UI agent, orchestrator, verifier) can declare and verify against. Eliminate the "implicit requirement" problem by forcing per-item user confirmation.
-
-## elicit
-
-elicit=true — this task REQUIRES user interaction. Never skip elicitation for efficiency.
-
-## Inputs
-
-- `{AppName}` argument (required for the canonical filename; ask once if missing).
-- `{topic}` argument (optional) — the product/feature area. If absent, ask once: "What's the topic / product area for this BRD?"
-- Existing BRD path. Resolve in this order: `core-config.yaml → customTechnicalDocuments`, then `docs/{AppName}-BRD.md` (legacy fallbacks: `docs/brd.md`, `docs/BRD.md`). If found, OPEN it and continue from the highest existing BRD-N. If not found, create `docs/{AppName}-BRD.md` from `.tfcore/templates/v4custom/app-brd-tmpl.md` (feature catalog + §10 BRD-N ledger shape — NOT ad-hoc "Phase N" sections).
-
-## SEQUENTIAL Execution
-
-### 1. Read or scaffold the BRD
-
-- If the resolved BRD exists: read it, parse all existing `BRD-{N}` IDs, and record the highest N. New requirements continue from N+1.
-- If absent: scaffold `docs/{AppName}-BRD.md` from `.tfcore/templates/v4custom/app-brd-tmpl.md` (leave unconfirmed sections empty; confirmed requirements land in the §10 ledger).
-- Ask the user which areas they want to scope today.
-
-### 2. Drive elicitation per area, one at a time
-
-For each area the user wants to cover (UI, data, workflow, NFRs, etc.):
-
-1. Generate a numbered list of CANDIDATE requirements for that area — 3 to 7 items, written as one-line "user can X" / "system shall Y" statements. Do NOT assign BRD-N IDs yet.
-2. Present the list and ask the user to (a) accept, (b) reject, (c) refine, or (d) add their own — one item at a time. Use this exact prompt format:
-   ```
-   Candidate requirements for {area}:
-   1. {candidate}
-   2. {candidate}
-   ...
-   For each, reply: keep / drop / refine: <new text>. Or add a new one with: add: <text>.
-   ```
-3. After the user processes each item, assign the next BRD-N to every KEPT or REFINED item, in the order they were confirmed. Show the assignments back:
-   ```
-   Assigned:
-   - BRD-{N}: {final text}
-   - BRD-{N+1}: {final text}
-   ```
-4. Append the assigned requirements to the §10 Functional-requirements ledger of the resolved BRD file (and flesh out / create the owning feature-catalog entry).
-
-Repeat for each area until the user says they're done with this BRD pass.
-
-### 3. NFRs and acceptance criteria (optional pass)
-
-If the user wants NFRs (performance, accessibility, browser support, etc.) or acceptance criteria for any BRD-N, run the same elicitation pattern. NFRs get their own BRD-N IDs (do NOT use a separate `NFR-N` series — keep one ID space so the verifier can grade everything uniformly).
-
-### 4. Final output
-
-Append/refresh the footer block of the BRD:
-```
----
-Last updated: {YYYY-MM-DD}
-Highest BRD ID: BRD-{N}
-```
-
-Print a compact summary to the user:
-```
-BRD updated → docs/{AppName}-BRD.md
-Added: BRD-{a}..BRD-{b} ({count} requirements)
-Phases touched: {list}
-```
-
-HALT.
-
-## Hard rules
-
-- elicit=true — never batch-confirm; one item at a time.
-- BRD-N IDs are append-only — never renumber an existing ID even if the user removes a requirement later (mark it `~~BRD-N~~ (removed)` instead).
-- Requirements are written as observable outcomes, not implementation hints. The verifier needs to be able to test them.
-- The analyst writes the BRD; the analyst does NOT implement anything.