reposition CaseForge as an implementation blueprint tool

Author: ADiamond12

CHANGELOG.md

@@ -6,13 +6,13 @@ Initial CaseForge Studio release.
 
 Highlights:
 
-- deterministic dossier pipeline
+- deterministic blueprint pipeline
 - CLI for create, list, show, and serve
 - local web UI
 - preview endpoint
-- recent dossier listing
+- recent blueprint listing
 - comparison-friendly saved-run summaries
-- multi-run dossier comparison through `/api/dossiers/compare`
+- multi-run blueprint comparison through `/api/dossiers/compare`
 - optional OpenAI provider overlay with graceful fallback
 - release checklist
 - packaging/install guidance for local and distributable use
@@ -22,4 +22,4 @@ Highlights:
 
 - live OpenAI happy-path validation with real credentials
 - comparison history beyond two runs
-- reviewer annotations on saved dossiers
+- annotation support on saved blueprints

README.md

@@ -1,22 +1,22 @@
 # CaseForge Studio
 
-CaseForge Studio turns a rough product idea into an interview-ready dossier through a deterministic multi-stage pipeline. It ships with a local web app, CLI, local HTTP API, persisted run history, and an optional OpenAI overlay that refines the final dossier without becoming the only path through the product.
+CaseForge Studio turns a rough product idea into an implementation-ready project blueprint through a deterministic multi-stage pipeline. It ships with a local web app, CLI, local HTTP API, persisted run history, and an optional OpenAI overlay that refines the final blueprint without becoming the only path through the product.
 
-The core thesis is simple: a strong portfolio project should be explainable, scoped, and reviewable. CaseForge Studio helps turn vague ideas into artifacts that are easier to demo, compare, and improve.
+The core thesis is simple: a strong project idea should become explainable, scoped, reviewable, and ready for execution. CaseForge Studio helps turn vague ideas into artifacts that are easier to evaluate, compare, and improve.
 
 ## Why It Matters
 
-- Deterministic by default, so the product is demoable without external services.
-- Optional live-provider overlay for a stronger AI story when credentials are available.
+- Deterministic by default, so the product remains usable without external services.
+- Optional live-provider overlay for stronger AI-assisted refinement when credentials are available.
 - Multiple surfaces from one shared service layer: CLI, browser UI, and HTTP API.
 - Saved runs and comparison views make iteration visible instead of hand-wavy.
-- Fast to explain in an interview: brief in, dossier out, compare runs, choose the strongest artifact.
+- Fast to evaluate in a technical review: brief in, blueprint out, compare runs, choose the strongest implementation path.
 
 ## Feature Set
 
-- Planner, architect, evaluator, and storyteller stages
+- Planner, architect, evaluator, and delivery-path stages
 - Markdown, JSON, and summary export under `outputs/` for persisted local runs
-- Local web app with dossier preview, saved-run browsing, and comparison
+- Local web app with blueprint preview, saved-run browsing, and comparison
 - Local HTTP API for generation, preview, retrieval, and compare flows
 - Optional OpenAI Responses API overlay with deterministic fallback
 - Standard-library-only backend runtime
@@ -30,12 +30,12 @@ Brief
   -> planner
   -> architect
   -> evaluator
-  -> storyteller
-  -> dossier export
+  -> delivery path
+  -> blueprint export
   -> optional OpenAI public-facing overlay
 ```
 
-The deterministic path is the primary product path. The live provider is an enhancement layer, not a dependency for the base demo.
+The deterministic path is the primary product path. The live provider is an enhancement layer, not a dependency for the base workflow.
 
 ## Quickstart
 
@@ -45,16 +45,16 @@ From the `caseforge-studio` project root:
 python -m pip install -e .
 ```
 
-Generate a saved dossier:
+Generate a saved blueprint:
 
 ```powershell
-python -m caseforge create "Build an AI interview coach that turns a resume and job description into practice questions, STAR prompts, and a confidence score."
+python -m caseforge create "Build an AI operations copilot that turns incident notes, service metrics, and follow-up tasks into a release-ready action plan."
 ```
 
-Preview a dossier without persistence:
+Preview a blueprint without persistence:
 
 ```powershell
-python -m caseforge create "Build an AI interview coach." --preset ml --preview --json
+python -m caseforge create "Build an AI operations copilot." --preset ml --preview --json
 ```
 
 Run the local web app:
@@ -85,7 +85,7 @@ The live provider path is optional. Without credentials, the app falls back to t
 ```powershell
 $env:OPENAI_API_KEY="your-key"
 $env:OPENAI_MODEL="gpt-5-mini"
-python -m caseforge create "Build an AI interview coach." --preset ml --provider openai --preview --json
+python -m caseforge create "Build an AI operations copilot." --preset ml --provider openai --preview --json
 ```
 
 Supported environment variables:
@@ -107,16 +107,16 @@ python -m caseforge create "Design an operations copilot that summarizes inciden
 Create from a file:
 
 ```powershell
-python -m caseforge create --brief-file examples/briefs/ai-interview-coach.md --goal "Show AI judgment"
+python -m caseforge create --brief-file examples/briefs/ai-ops-copilot.md --mode "AI workflow product" --goal "Emphasize shipping discipline" --preset full-stack
 ```
 
-List recent dossiers:
+List recent blueprints:
 
 ```powershell
 python -m caseforge list
 ```
 
-Show a persisted record:
+Open a persisted record:
 
 ```powershell
 python -m caseforge show <slug>
@@ -136,22 +136,22 @@ Example request:
 
 ```json
 {
-  "brief": "Build an AI interview coach that turns a resume and job description into mock questions and rehearsal plans.",
-  "audience": "Hiring manager",
+  "brief": "Build an AI operations copilot that turns incident notes, service metrics, and follow-up tasks into a release-ready action plan.",
+  "audience": "Technical stakeholders",
   "mode": "AI assistant",
-  "goal": "Show AI judgment",
+  "goal": "Emphasize AI decisioning",
   "preset": "ml",
   "provider": "openai",
   "provider_model": "gpt-5-mini"
 }
 ```
 
-## Demo Flow
+## Usage Flow
 
 1. Start the web app and paste a rough project idea into the brief box.
-2. Generate a dossier and show the score, architecture section, and interview story.
-3. Open the committed sample dossier at `examples/sample-dossier.md` or a locally generated artifact under `outputs/<slug>/dossier.md`.
-4. Compare two runs to show score movement, provider path, and recommendation changes.
+2. Generate a blueprint and review the score, architecture section, and delivery path.
+3. Open the committed sample blueprint at `examples/sample-blueprint.md` or a locally generated artifact under `outputs/<slug>/dossier.md`.
+4. Compare two runs to review score movement, provider path, and recommendation changes.
 5. Explain why the deterministic path is the default and when the live provider is worth using.
 6. Close with the test suite and release checklist.
 
@@ -161,8 +161,8 @@ Example request:
 - `python -m build`
 - smoke-test `GET /health`
 - smoke-test `GET /`
-- generate one dossier through the CLI
-- generate one dossier through the web UI
+- generate one blueprint through the CLI
+- generate one blueprint through the web UI
 
 ## Project Layout
 
@@ -171,15 +171,15 @@ caseforge-studio/
 |-- .github/
 |-- caseforge/
 |-- examples/
-|   |-- sample-dossier.md
+|   |-- sample-blueprint.md
 |   `-- briefs/
 |-- tests/
 |-- CHANGELOG.md
 |-- RELEASE_CHECKLIST.md
 `-- README.md
 ```
 
-`outputs/` is generated locally at runtime and is intentionally kept out of version control. The committed public sample dossier lives at `examples/sample-dossier.md`.
+`outputs/` is generated locally at runtime and is intentionally kept out of version control. The committed public sample blueprint lives at `examples/sample-blueprint.md`.
 
 ## Release And Security
 
@@ -192,12 +192,12 @@ caseforge-studio/
 
 - The server is local-first and single-tenant.
 - There is no built-in authentication or multi-user access control.
-- The OpenAI overlay path should be demonstrated only with deliberate credential handling.
+- The OpenAI overlay path should be used only with deliberate credential handling.
 - The default runtime is still a local stdlib HTTP server, not a multi-tenant hosted deployment stack.
 
 ## Next Steps
 
 - Add a clean deployment wrapper around the local server path
-- Capture one intentional live-provider demo artifact
+- Capture one intentional live-provider blueprint artifact
 - Add stronger browser-level regression coverage
-- Promote the best saved dossier flow into a tighter public release
+- Promote the best saved blueprint flow into a tighter public release

RELEASE_CHECKLIST.md

@@ -1,44 +1,44 @@
 # Release Checklist
 
-Use this before calling a CaseForge Studio version ready to demo or share.
+Use this before calling a CaseForge Studio version ready to use or share.
 
 ## Product
 
 - The web app loads without console-breaking behavior.
-- `POST /api/dossiers` works and persists a dossier.
+- `POST /api/dossiers` works and persists a blueprint artifact.
 - `POST /api/dossiers/preview` works without persistence.
 - `GET /api/dossiers` returns recent saved runs.
 - `GET /api/dossiers/compare` compares two saved runs.
-- `GET /api/dossiers/<slug>` returns a loadable dossier payload.
+- `GET /api/dossiers/<slug>` returns a loadable blueprint payload.
 - If `provider=openai` is requested without credentials, the app falls back cleanly.
 - If OpenAI credentials are configured, the live overlay path is explicitly tested.
-- If a public sample dossier is included, `examples/sample-dossier.md` is current and readable.
+- If a public sample blueprint is included, `examples/sample-blueprint.md` is current and readable.
 
 ## Verification
 
 - Run `python -m unittest discover -s tests -v`
 - Smoke-test `GET /health`
 - Smoke-test `GET /`
-- Generate one dossier from the CLI
-- Generate one dossier from the UI
+- Generate one blueprint from the CLI
+- Generate one blueprint from the UI
 
 ## Documentation
 
 - `README.md` reflects the current commands and endpoints.
 - `CHANGELOG.md` reflects completed unreleased work.
-- If a curated sample dossier is included for public demo, it is current and sanitized.
+- If a curated sample blueprint is included for public sharing, it is current and sanitized.
 - Any new behavior is described briefly and concretely.
 
-## Interview Readiness
+## Industry Readiness
 
-- The deterministic fallback story is still true.
-- The live product path is easy to explain in under two minutes.
+- The deterministic fallback and local-first story are still true.
+- The live product path is easy to explain in a quick handoff.
 - The strongest tradeoff is explicit.
 - The sample prompt and sample output both feel intentional.
 
 ## Cleanup
 
-- Remove throwaway outputs that do not support the demo story.
-- Keep one strong sample dossier.
+- Remove throwaway outputs that do not support the current release story.
+- Keep one strong sample blueprint.
 - Do not leave stale claims about endpoints or features in docs or generated artifacts.
 - Confirm `.github/workflows/ci.yml` and `.github/workflows/release.yml` still match the packaging story.

SECURITY.md

@@ -2,7 +2,7 @@
 
 ## Scope
 
-CaseForge Studio is currently a local-first, single-tenant application. It is designed for workstation demos, product iteration, and portfolio presentation, not for untrusted multi-tenant deployment in its current form.
+CaseForge Studio is currently a local-first, single-tenant application. It is designed for local validation, product iteration, and technical review, not for untrusted multi-tenant deployment in its current form.
 
 ## Supported Surface
 
@@ -13,8 +13,8 @@ CaseForge Studio is currently a local-first, single-tenant application. It is de
 ## Security Expectations
 
 - Do not commit API keys, tokens, or `.env` files.
-- Treat persisted dossier content as local project data unless you intentionally sanitize it for sharing.
-- Use the deterministic provider path by default when recording demos or testing on untrusted environments.
+- Treat persisted blueprint content as local project data unless you intentionally sanitize it for sharing.
+- Use the deterministic provider path by default when capturing walkthroughs or testing on untrusted environments.
 - Do not expose the local server directly to the public internet without adding authentication, reverse-proxy controls, and deployment hardening.
 
 ## Reporting

caseforge/cli.py

@@ -12,41 +12,41 @@
 def build_parser() -> argparse.ArgumentParser:
     parser = argparse.ArgumentParser(
         prog="caseforge",
-        description="Generate interview-ready project dossiers from rough ideas.",
+        description="Generate implementation-ready project blueprints from rough ideas.",
     )
     subparsers = parser.add_subparsers(dest="command", required=True)
 
-    create_parser = subparsers.add_parser("create", help="generate a dossier from a brief")
+    create_parser = subparsers.add_parser("create", help="generate a project blueprint from a brief")
     create_parser.add_argument("brief", nargs="?", help="raw project brief text")
     create_parser.add_argument("--brief-file", help="path to a markdown or text brief")
     create_parser.add_argument("--title")
-    create_parser.add_argument("--audience", default="Hiring manager")
+    create_parser.add_argument("--audience", default="Technical stakeholders")
     create_parser.add_argument("--mode", default="AI assistant")
-    create_parser.add_argument("--goal", default="Show systems thinking")
+    create_parser.add_argument("--goal", default="Drive implementation clarity")
     create_parser.add_argument(
         "--preset",
         default="general",
         choices=["general", "product", "ml", "full-stack", "founder"],
-        help="evaluation preset for the dossier",
+        help="evaluation preset for the generated blueprint",
     )
     create_parser.add_argument(
         "--provider",
         default="deterministic",
         choices=["deterministic", "openai"],
-        help="generation provider; deterministic is the default safe demo path",
+        help="generation provider; deterministic is the default safe local path",
     )
     create_parser.add_argument("--provider-model", help="optional live provider model override")
-    create_parser.add_argument("--preview", action="store_true", help="generate a dossier without saving it")
+    create_parser.add_argument("--preview", action="store_true", help="generate a blueprint without saving it")
     create_parser.add_argument("--json", action="store_true", help="print the public JSON payload")
 
     serve_parser = subparsers.add_parser("serve", help="run the local web app")
     serve_parser.add_argument("--host", default="127.0.0.1")
     serve_parser.add_argument("--port", type=int, default=8127)
 
-    show_parser = subparsers.add_parser("show", help="show a persisted dossier record by slug")
+    show_parser = subparsers.add_parser("show", help="open a persisted blueprint record by slug")
     show_parser.add_argument("slug")
 
-    list_parser = subparsers.add_parser("list", help="list recent dossiers")
+    list_parser = subparsers.add_parser("list", help="list recent blueprints")
     list_parser.add_argument("--limit", type=int, default=10)
 
     return parser
@@ -100,7 +100,7 @@ def main(argv: list[str] | None = None) -> int:
     if args.command == "list":
         records = service.list_records(limit=args.limit)
         if not records:
-            print("No dossiers found.")
+            print("No blueprints found.")
             return 0
         for record in records:
             print(f"{record['slug']}\t{record['title']}\t{record['score']}")

caseforge/markdown.py

@@ -82,15 +82,15 @@ def render_markdown(
         "### Mitigations",
         *[f"- {item}" for item in evaluator.mitigations],
         "",
-        "## Interview Pitch",
+        "## Execution Briefing",
         storyteller.elevator_pitch,
         "",
-        f"Hook: {storyteller.interviewer_hook}",
+        f"Delivery hook: {storyteller.delivery_hook}",
         "",
-        "### Demo Script",
+        "### Execution Walkthrough",
         *[f"- {item}" for item in storyteller.demo_script],
         "",
-        "### Talking Points",
+        "### Key Points",
         *[f"- {item}" for item in storyteller.talking_points],
         "",
     ]

caseforge/models.py

@@ -9,9 +9,9 @@
 class ProjectBrief:
     brief: str
     title: str | None = None
-    audience: str = "interview"
+    audience: str = "Technical stakeholders"
     mode: str = "AI assistant"
-    goal: str = "Show systems thinking"
+    goal: str = "Drive implementation clarity"
     preset: str = "general"
     provider: str = "deterministic"
     provider_model: str | None = None
@@ -51,7 +51,7 @@ class EvaluatorResult:
 @dataclass(frozen=True, slots=True)
 class StorytellerResult:
     elevator_pitch: str
-    interviewer_hook: str
+    delivery_hook: str
     demo_script: tuple[str, ...]
     talking_points: tuple[str, ...]
 
@@ -161,10 +161,10 @@ def sections(self) -> tuple[DossierSection, ...]:
             ),
             DossierSection(
                 label="Audience",
-                title="Who the demo is aimed at",
+                title="Who this blueprint is aimed at",
                 body=(
-                    f"The dossier is optimized for {self.brief.audience.lower()} review and tuned to "
-                    f"show {self.brief.goal.lower()} through a {self.brief.mode.lower()} lens."
+                    f"The blueprint is optimized for {self.brief.audience.lower()} alignment, uses "
+                    f"{self.brief.mode.lower()} framing, and is designed to {self.brief.goal.lower()}."
                 ),
             ),
             DossierSection(
@@ -186,9 +186,9 @@ def sections(self) -> tuple[DossierSection, ...]:
                 ),
             ),
             DossierSection(
-                label="Interview story",
-                title="How to present it",
-                body=self.storyteller.interviewer_hook,
+                label="Delivery story",
+                title="How to move it toward implementation",
+                body=self.storyteller.delivery_hook,
             ),
         )