stackprep-pro 0.2.4__tar.gz → 0.2.6__tar.gz

stackprep_pro-0.2.6/CLAUDE.md

@@ -0,0 +1,42 @@
+# CLAUDE.md — stackprep-pro
+
+Rules for working on this repo. **Do NOT violate these. Do NOT touch things that already work.**
+
+## Golden rule
+
+- **If something works, do NOT change it.** No "improvements", no refactors, no edits to working code. Only touch what is explicitly broken and being fixed.
+- Always **follow the skill files** in `src/stackprep_pro/skills/`. The skill is the source of truth for behavior. Never invent a flow that contradicts the skill.
+
+## Startup / session flow (MUST follow exactly)
+
+1. When the MCP starts, **ask the user in plain human language**: "Are you prepping for a technical interview or a certification exam?"
+2. **Only after** the user picks interview or certification:
+   - If saved sessions exist for that mode, offer to **continue a saved session** (shown by the **name the user gave it**) or **start a new one**.
+   - If none exist, just start a new session.
+3. **Never expose backend tool names** to the user (no "list_sessions", "start_session", "list_study_packs", etc.). Use natural language only.
+4. **Never auto-call** `list_sessions` or `list_study_packs` on startup. Call them silently in the background **only** when the user explicitly asks to continue/load.
+
+## Certification exam version
+
+- **Never hardcode exam versions** (e.g. do NOT hardcode "SnowPro Core = COF-C03"). It must work for **ANY** certification.
+- Always pull the **latest exam version** dynamically (web search / latest official exam guide), exactly as the original skill does.
+- **`cert_name` must be passed exactly as the user typed it.** Never modify, correct, or substitute it from training data (e.g. do NOT turn COF-C03 into COF-C02).
+
+## Study pack
+
+- The user **must be able to name** the study pack when saving it.
+- Study pack format is **markdown, NOT JSON**. For each topic: official docs link (or community source like Reddit/Discord/LinkedIn if no official docs exist — e.g. niche certs like NVIDIA Agentic AI), best video resource, and a 2–3 sentence summary. Include percentages/scores and links as the original skill does.
+- Never fabricate documentation URLs. Only real, publicly accessible links.
+
+## Versioning & publishing
+
+- This package publishes to **PyPI** as `stackprep-pro` via **GitHub Actions** on push to `main` (`.github/workflows/publish.yml`).
+- **The human bumps the version in `pyproject.toml` and commits it.** The workflow does **NOT** auto-commit or auto-bump — it only reads the version, tags it, and publishes.
+- Every code change that ships needs a **version bump** (PyPI rejects re-publishing an existing version).
+- After publishing, to test the new version: `uv cache clean`, then restart the MCP client (uvx caches old versions).
+- `git pull` is set to rebase in this repo. If branches diverge, `git pull --rebase`.
+
+## Commit messages
+
+- **ALWAYS give a commit title after making any change.** Every time you edit files, end your response with a single concise conventional-commit title — never skip it, never make the user ask for it.
+- Do not commit/push unless explicitly asked.

{stackprep_pro-0.2.4 → stackprep_pro-0.2.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: stackprep-pro
-Version: 0.2.4
+Version: 0.2.6
 Summary: stackprep-pro — interview & certification prep MCP server for any AI client
 Project-URL: Homepage, https://github.com/youngpada1/stackprep-pro
 Project-URL: Repository, https://github.com/youngpada1/stackprep-pro
@@ -168,7 +168,7 @@ Point this at any Dropbox, Google Drive, or OneDrive folder for cross-platform s
 | `save_study_pack` | Save the study pack content to disk. | `session_id`, `name`, `content` |
 | `list_sessions` | List all saved sessions. Call this silently in the background only when the user says they want to continue a previous session. Never mention this tool to the user. |  |
 | `resume_session` | Resume a previously saved session. Returns full session state and skill rules. | `session_id` |
-| `list_study_packs` | List all saved study packs. |  |
+| `list_study_packs` | List all saved study packs. Call this silently only when the user explicitly asks to see or load a saved study pack. Never call this on startup or automatically. Never mention this tool to the user. |  |
 | `load_study_pack` | Load a previously saved study pack by name. | `name` |
 
 ---

{stackprep_pro-0.2.4 → stackprep_pro-0.2.6}/README.md

@@ -148,7 +148,7 @@ Point this at any Dropbox, Google Drive, or OneDrive folder for cross-platform s
 | `save_study_pack` | Save the study pack content to disk. | `session_id`, `name`, `content` |
 | `list_sessions` | List all saved sessions. Call this silently in the background only when the user says they want to continue a previous session. Never mention this tool to the user. |  |
 | `resume_session` | Resume a previously saved session. Returns full session state and skill rules. | `session_id` |
-| `list_study_packs` | List all saved study packs. |  |
+| `list_study_packs` | List all saved study packs. Call this silently only when the user explicitly asks to see or load a saved study pack. Never call this on startup or automatically. Never mention this tool to the user. |  |
 | `load_study_pack` | Load a previously saved study pack by name. | `name` |
 
 ---

{stackprep_pro-0.2.4 → stackprep_pro-0.2.6}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "stackprep-pro"
-version = "0.2.4"
+version = "0.2.6"
 description = "stackprep-pro — interview & certification prep MCP server for any AI client"
 readme = "README.md"
 license = { text = "MIT" }

{stackprep_pro-0.2.4 → stackprep_pro-0.2.6}/src/stackprep_pro/server.py

@@ -79,6 +79,14 @@ def start_session(
 ) -> str:
     """Start a new stackprep session. Returns a session ID and the skill rules for the AI to follow.
 
+    STARTUP FLOW (follow exactly, in plain language — never show tool or field names to the user):
+    1. First ask the user, in plain language, what they want to prep for (e.g. "What would you like to prep for?") and let them indicate interview or certification.
+    2. After they choose, check for saved sessions of that mode (call list_sessions silently).
+       - If matching saved sessions exist, ask: "Do you want to continue a saved session or start a new one?"
+         and list the saved sessions by the name the user gave them.
+       - If none exist, just proceed to start a new session.
+    3. For a new session, collect the inputs the skill requires, then call start_session.
+
     Args:
         mode: "interview" or "certification"
         cert_name: For certification mode — the exam name exactly as the user typed it (e.g. "AWS SAA-C03"). NEVER modify, correct, or substitute the cert name — use the user's exact input verbatim.
@@ -343,7 +351,7 @@ def resume_session(session_id: str) -> str:
 
 @mcp.tool()
 def list_study_packs() -> str:
-    """List all saved study packs."""
+    """List all saved study packs. Call this silently only when the user explicitly asks to see or load a saved study pack. Never call this on startup or automatically. Never mention this tool to the user."""
     packs = _packs_dir()
     files = sorted(packs.glob("*.json"))
     if not files:

{stackprep_pro-0.2.4 → stackprep_pro-0.2.6}/src/stackprep_pro/skills/certification.md

@@ -9,11 +9,33 @@ Adaptive certification exam prep — one question at a time, with instant feedba
 
 ## Session setup
 
-Inputs arrive via MCP (certification name, extra topics). Give a short 2–3 line summary:
-- Exam structure overview
-- Key domains and weightings from the latest official exam guide
+Inputs arrive via MCP (certification name, extra topics). After looking up the latest official exam guide, present a clean, structured overview in this exact layout:
 
-Then wait — questions are requested one at a time via `next_question`.
+```
+[Cert name + code] — confirmed active as of [date] (replaces [prev version] if applicable).
+
+**Exam structure:**
+— [N] questions | [time] min | Passing: [score] | [price] via [provider]
+
+**Domains & weightings:**
+
+| Domain                                   | Weight |
+|------------------------------------------|--------|
+| [Domain 1]                               | [X]%   |
+| [Domain 2]                               | [X]%   |
+| ...                                      | ...    |
+
+**Notable [version] additions:** [short note on what's new, if applicable].
+
+**Sources:**
+— [Title] ([url])
+— [Title] ([url])
+
+---
+Ready when you are — first question?
+```
+
+Then wait — questions are requested one at a time.
 
 ## Question format
 

{stackprep_pro-0.2.4 → stackprep_pro-0.2.6}/src/stackprep_pro/skills/interview.md

@@ -9,12 +9,29 @@ Adaptive technical interview prep — one question at a time, with instant feedb
 
 ## Session setup
 
-Inputs arrive via MCP (CV, job description, extra topics). Give a short 2–3 line summary:
-- Seniority level inferred from CV
-- Key domains
-- Top skill gaps vs. the job description
+Inputs arrive via MCP (CV, job description, extra topics). After analysing the CV and JD, present a clean, structured overview in this exact layout:
 
-Then wait — questions are requested one at a time via `next_question`.
+```
+[Role / target position] — [seniority level inferred from CV].
+
+**Interview focus:**
+— Based on your CV vs. the job description
+
+**Domains & focus:**
+
+| Domain                                   | Focus  |
+|------------------------------------------|--------|
+| [Domain 1]                               | [High/Med/Low] |
+| [Domain 2]                               | [High/Med/Low] |
+| ...                                      | ...    |
+
+**Top skill gaps vs. the JD:** [short note on the biggest gaps to drill].
+
+---
+Ready when you are — first question?
+```
+
+Then wait — questions are requested one at a time.
 
 ## Question format
 

{stackprep_pro-0.2.4 → stackprep_pro-0.2.6}/uv.lock

@@ -723,7 +723,7 @@ wheels = [
 
 [[package]]
 name = "stackprep-pro"
-version = "0.2.4"
+version = "0.2.5"
 source = { editable = "." }
 dependencies = [
     { name = "mcp", extra = ["cli"] },