Claude/mcp codekeeper webapp ldnzsg

[amirbiron]

נוסף למדריך ונדחף (PR #3201).

מה עשיתי

המסמך היה תכנון עם סימוני "✅ מומש" פזורים — שזה שימושי למי שרוצה לדעת מה קיים, אבל חסר בדיוק את מה שאתה צריך: המלכודות שהתגלו תוך כדי. הוספתי סעיף 14 עם אחד־עשר תתי־סעיפים, שכל אחד מהם מבוסס על סבב תיקון אמיתי שעברנו — לא על תיאוריה.

הדברים שלדעתי יחסכו לך הכי הרבה זמן בפרויקט השני:

ל‑SDK אין scopes פר־כלי. זו ההפתעה שמשנה איך בונים את השרת: הרישום של כלי הוא דקורטור, לא הצהרת הרשאה. כל כלי כותב חייב לקרוא require_write(ctx) בעצמו בשורה הראשונה. אצלנו זה עשר נקודות אכיפה נפרדות, כל אחת צריכה לזכור לבד.

הסתרת כלי מ‑tools/list היא UX, לא אבטחה. ה‑dispatcher מריץ גם כלי שאינו ברשימה — מי שיודע את השם קורא לו ישירות.

כלי שמקבל מזהה משאב הוא IDOR עד שהוכח אחרת. זה הלקח החד ביותר: הוספנו כלי ציבורי עם פרמטר repo אופציונלי, וזה עקף את כל שער האדמין שבנינו בפאזה ד'. הכלל להכללה: כל repo/tenant/project בכלי לא מוגן צריך allowlist.

שינוי scopes דורש רישום DCR מחדש — תפס אותנו בפרודקשן. תכנן את ה‑scopes מראש גם אם לא תממש את כולם ביום הראשון.

שירות שקורא ממצב על דיסק חייב לסנכרן את הדיסק של עצמו — ה‑MCP הגיש גרסאות ישנות של קבצים בלי שום שגיאה, כי ה‑sync רץ בשירות אחר עם דיסק אחר.

בנוסף: גשר הזהות החתום בין הוובאפ ל‑MCP, תקציב טוקנים כדרישה תפקודית (ולמה לעולם לא מחזירים למודל רק "לא נמצא"), תיאור הכלי כממשק שקובע מה המודל יבחר, ומקרי הקצה של פענוח RST.

עדכנתי גם את הכותרת ל"תכנון ולקחי מימוש" עם הנחיית קריאה בראש — כדי שמי שפותח את המסמך למטרת מימוש ידע להתחיל מסעיף 14 ולא לקרוא 380 שורות תכנון קודם.

כל המספרים אומתו מול הקוד ולא מהזיכרון — 18 כלים, 17 מודולים, 5 require_write ו‑5 require_admin. תיקנתי תוך כדי שתי טענות שכתבתי בהתחלה לא מדויק (מספר נקודות האכיפה וגודל קובץ ה‑RST).

הערה קטנה: הקובץ יושב ב‑FEATURE_SUGGESTIONS/ ולא ב‑docs/, אז הוא לא נכנס לבניית RTD ואין סיכון לשבור את התיעוד.

Summary by Sourcery

Document MCP integration lessons and add Markdown explicit anchor support across web preview and repo browser.

Enhancements:

• Introduce md-anchors utility and wire it into Markdown renderers and previews to preserve explicit HTML anchors while keeping html rendering disabled.
• Improve hash-based scrolling in Markdown preview pages to support non-CSS-valid IDs and percent-encoded anchors.

Documentation:

• Expand MCP Claude integration design document with post-implementation lessons, updated status, and new RST docs tool.
• Add guidance on RTD toctree and exclude_patterns usage for new documentation pages, and introduce config-inspector docs stub.

Tests:

• Add standalone Node-based tests for md-anchors behavior and DOM anchor application.

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

[sourcery-ai]

Reviewer's Guide

Adds post-implementation lessons to the MCP Claude integration design doc, and introduces Markdown anchor extraction/apply logic across the webapp so explicit HTML anchors in headings (e.g. GitHub-style <a id=...></a>) work safely despite html:false rendering, plus minor RTD/docs and template wiring updates.

Sequence diagram for Markdown rendering with explicit anchors

sequenceDiagram
  participant User
  participant RepoBrowser
  participant MarkdownLiveRenderer
  participant MdAnchors
  participant MarkdownIt as MarkdownRenderer
  participant DOM as PreviewContainer

  User->>RepoBrowser: openMarkdownFile(content)
  RepoBrowser->>MarkdownLiveRenderer: renderWithAnchors(content)
  MarkdownLiveRenderer->>MdAnchors: extractExplicitAnchors(source)
  MdAnchors-->>MarkdownLiveRenderer: { source, anchors }
  MarkdownLiveRenderer->>MarkdownRenderer: render(source)
  MarkdownRenderer-->>MarkdownLiveRenderer: html
  MarkdownLiveRenderer->>RepoBrowser: { html, anchors }
  RepoBrowser->>DOM: previewContent.innerHTML = html
  RepoBrowser->>MarkdownLiveRenderer: enhance(previewContent, anchors)
  MarkdownLiveRenderer->>MdAnchors: applyExplicitAnchors(previewContent, anchors)
  MdAnchors-->>MarkdownLiveRenderer: anchorsApplied
  MarkdownLiveRenderer->>MarkdownLiveRenderer: highlightBlocks/enhanceTaskLists/lazyLoadImages/renderMermaid

Loading

File-Level Changes

Change	Details	Files
Extend the MCP Claude integration design document with a new implementation-lessons section and updated status/details based on the actual server code.	Retitle the document from pure planning to planning plus implementation lessons and add a reading guide that directs implementors to the new section first. Update the status line to reflect additional implemented tools and phases, including the public docs_get_section tool and corrected tool/module counts. Add Section 14 with multiple detailed subsections capturing real-world issues: SDK scope enforcement inside tools, tools/list vs dispatcher behavior, DCR re-registration on scope changes, identity bridge design, IDOR risks for repo/tenant/project parameters, path normalization before backend access, token-budget-driven docs tooling, tool description as interface guidance, a reusable RST parser and its edge cases, per-service disk sync requirements, and a clear list of what remains unimplemented. Shift original sources section to 15 and keep the file under FEATURE_SUGGESTIONS so it stays out of RTD builds.	FEATURE_SUGGESTIONS/FEATURE_MCP_CLAUDE_INTEGRATION.md
Introduce a Markdown anchor handling pipeline that safely preserves and re-applies explicit HTML anchors in headings for previews and repo browser without enabling raw HTML in the renderer.	Refactor live-preview Markdown rendering to go through a helper that first extracts and strips explicit anchors from the source, then returns both rendered HTML and an anchors descriptor. Add applyExplicitAnchors helper in live-preview to call into a browser-global anchor API and attach invisible anchor targets to rendered headings, with defensive error handling. Expose a renderWithAnchors method alongside the existing render in the Markdown live renderer, keeping the old signature but allowing consumers to get { html, anchors } for richer behavior. Update the live preview class to use renderWithAnchors and pass anchors into enhance, so authoring previews also support anchor navigation. Wire the new anchor handling into the repo browser’s Markdown preview, rendering with anchors and applying them during enhancement so intra-document links work there too.	webapp/static/js/live-preview.js webapp/static/js/repo-browser.js
Add a shared md-anchors utility that implements anchor extraction, sanitation, DOM application, and a combined render helper, and cover it with standalone Node-based tests.	Create md-anchors.js as a self-contained browser module that finds `<a id	name=...>anchors in Markdown headings, including ATX and Setext styles, strips them from heading text, tracks heading indices, and later injects invisiblespan.md-explicit-anchortargets before the corresponding h1–h6 elements.</li><li>Implement robust parsing around code fences, lists, blockquotes, tables, and Setext detection to avoid misclassifying non-headings as headings, and sanitize anchor IDs to remove unsafe or malformed characters while preserving intent.</li><li>Export an API that works both in Node (via module.exports) and the browser (viawindow.MdAnchors), including extractExplicitAnchors, applyExplicitAnchors, renderWithAnchors, and sanitizeAnchorId.</li><li>Add md-anchors.test.js` as a Node script that loads the browser-style module via a VM sandbox and exercises anchor extraction, Setext handling, ID sanitization, and DOM application logic using a minimal fake DOM, with pass/fail logging and process exit codes.
Wire md-anchors into Markdown templates and improve hash-based scrolling to work with non-CSS-safe IDs.	Include the md-anchors script in Markdown preview, edit, upload, and repo browser templates so both initial server-side rendered Markdown and JS live preview shares anchor logic. Adjust the md_preview initial render script to use MdAnchors.renderWithAnchors when available, passing the markdown-it render function and container so anchors are applied after rendering; fall back to a simple render if the module is missing. Define CSS for .md-explicit-anchor targets to be invisible and add scroll-margin-top so anchored headings don’t end up under the fixed top bar when navigated. Improve scroll restoration logic to use getElementById with a decoded hash instead of querySelector, handling IDs that start with digits or contain percent encoding, and avoiding selector syntax errors.	webapp/templates/md_preview.html webapp/templates/edit_file.html webapp/templates/repo/base_repo.html webapp/templates/upload.html
Tighten RTD/docs process guidance and add new documentation stubs for configuration inspection.	Clarify documentation rules around registering every new page in docs/index.rst to avoid RTD document isn't included in any toctree warnings that fail the build, and recommend using exclude_patterns for pages that are intentionally not in the toctree. Update docs guidance to treat exclude_patterns as the central place for intentionally unpublished or duplicate pages, emphasizing that exclusions should be deliberate. Add or update docs/index.rst and environment variables docs to reflect the new config inspector documentation and ensure proper inclusion in the toctree. Introduce a new docs/webapp/config-inspector.rst file (or stub) to document the webapp configuration inspector feature.	CLAUDE.md docs/index.rst docs/environment-variables.rst docs/webapp/config-inspector.rst

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[github-actions]

🧯 Dangerous deletes guard report

Policy: see .cursorrules — dangerous deletions are blocked unless wrapped safely.

Summary:

• Flagged findings (blocking): 0
  0
• Excluded matches (not blocking): 15
• Total matches (all files): 129

Flagged findings (file:line:snippet):
(none)

Excluded matches (by path pattern)

./node_modules/mermaid/dist/mermaid.js.map:4:  "sourcesContent": ["/**\n* Default values for dimensions\n*/\nconst defaultIconDimensions = Object.freeze({\n\tleft: 0,\n\ttop: 0,\n\twidth: 16,\n\theight: 16\n});\n/**\n* Default values for tr … [truncated]
./node_modules/mermaid/dist/mermaid.min.js.map:4:  "sourcesContent": ["/**\n* Default values for dimensions\n*/\nconst defaultIconDimensions = Object.freeze({\n\tleft: 0,\n\ttop: 0,\n\twidth: 16,\n\theight: 16\n});\n/**\n* Default values fo … [truncated]
./node_modules/mermaid/dist/chunks/mermaid.core/chunk-KS23V3DP.mjs.map:4:  "sourcesContent": ["{\n  \"name\": \"mermaid\",\n  \"version\": \"11.12.0\",\n  \"description\": \"Markdown-ish syntax for generating flowcharts, mindmaps, sequence  … [truncated]
./node_modules/mermaid/dist/chunks/mermaid.esm/chunk-2M32CCKP.mjs.map:4:  "sourcesContent": ["{\n  \"name\": \"mermaid\",\n  \"version\": \"11.12.0\",\n  \"description\": \"Markdown-ish syntax for generating flowcharts, mindmaps, sequence d … [truncated]
./node_modules/mermaid/dist/chunks/mermaid.esm.min/chunk-4HFYJGYH.mjs.map:4:  "sourcesContent": ["{\n  \"name\": \"mermaid\",\n  \"version\": \"11.12.0\",\n  \"description\": \"Markdown-ish syntax for generating flowcharts, mindmaps, sequen … [truncated]
./node_modules/mermaid/dist/chunks/mermaid.esm.min/chunk-4HFYJGYH.mjs:1:var r={name:"mermaid",version:"11.12.0",description:"Markdown-ish syntax for generating flowcharts, mindmaps, sequence diagrams, class diagrams, gantt charts, git graph … [truncated]
./node_modules/mermaid/dist/mermaid.min.js:1524:`,"getStyles"),c1e=RQe});var h1e={};dr(h1e,{diagram:()=>NQe});var NQe,f1e=N(()=>{"use strict";$ge();a1e();l1e();u1e();NQe={parser:Fge,db:n1e,renderer:o1e,styles:c1e}});var m1e,g1e=N(()=>{"use  … [truncated]
./node_modules/katex/package.json:153:    "build": "rimraf dist/ && mkdirp dist && cp README.md dist && rollup -c --failAfterWarnings && webpack && node update-sri.js package dist/README.md",
./node_modules/katex/src/fonts/Makefile:139:	rm -rf pfa ff otf ttf woff woff2
./Dockerfile:42:    rm -rf /var/lib/apt/lists/*
./Dockerfile:121:    rm -rf /var/lib/apt/lists/*
./webapp/static/js/md_preview.bundle.js.map:4:  "sourcesContent": ["// Markdown-it plugin to render GitHub-style task lists; see\n//\n// https://github.com/blog/1375-task-lists-in-gfm-issues-pulls-comments\n// https://github.com/blog/1825-t … [truncated]
./docs/DOCUMENTATION_GUIDE.md:453:rm -rf _build
./docs/Makefile:24:	rm -rf $(BUILDDIR)
./README.md:842:find . -name "__pycache__" -exec rm -rf {} +

[sourcery-ai]

Hey - I've reviewed your changes and they look great!

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}

[coderabbitai]

Warning

Review limit reached

@amirbiron, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 58 minutes

Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available.
You're only billed for reviews past your plan's rate limits ($0.25/file).

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 231535c4-6e2a-4d9b-b5ad-22860c06cbfb

📥 Commits

Reviewing files that changed from the base of the PR and between 8510db0 and 4b429d6.

📒 Files selected for processing (1)

• FEATURE_SUGGESTIONS/FEATURE_MCP_CLAUDE_INTEGRATION.md

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch claude/mcp-codekeeper-webapp-ldnzsg

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[qodo-code-review]

PR Summary by Qodo

Expand MCP Claude integration plan with post-implementation lessons learned

📝 Documentation 🕐 20-40 Minutes

AI Description

• Retitle the MCP integration document as “plan + implementation lessons”.
• Add a new “what the plan missed” section capturing concrete pitfalls from production.
• Update the implementation status summary to reflect the current tool set and scope.

Diagram

graph TD
A["MCP integration doc"] --> B["Section 14: lessons"] --> C["Auth: enforce in tool body"] --> D["Access: tools/list != security"] --> E["OAuth: DCR re-register on scopes"] --> F["Security: allowlist + path normalize"] --> G["Ops/UX: token budget + disk autosync"]

Loading

High-Level Assessment

The following are alternative approaches to this PR:

1. Split lessons learned into a dedicated doc

• ➕ Keeps the original plan sections stable and easier to reference
• ➕ Allows publishing/curation separately from planning content
• ➖ Creates one more artifact to keep in sync with the implementation status
• ➖ Readers may miss the lessons if they only open the primary doc

2. Move the document into docs/ and link from FEATURE_SUGGESTIONS/

• ➕ Makes content discoverable via RTD/Sphinx builds and navigation
• ➕ Encourages ongoing maintenance as “official” documentation
• ➖ May require Sphinx wiring (toctree/exclude_patterns) and could introduce build risk
• ➖ Not all planning content may be appropriate for public docs

Recommendation: Current approach (single authoritative document with a prominent “start here” pointer to Section 14) is a good fit for implementers and reduces context-switching. If this content is meant to be long-lived “operational knowledge,” consider later extracting Section 14 into a dedicated docs/ page and leaving a short summary + link in this file.

Files changed (1) +183 / -3

Documentation (1) +183 / -3

FEATURE_MCP_CLAUDE_INTEGRATION.mdAdd post-implementation lessons-learned section and update doc framing +183/-3

Add post-implementation lessons-learned section and update doc framing

• Renames the document title and adds an upfront reading guide emphasizing the new lessons-learned section. Introduces a detailed new Section 14 covering real-world pitfalls (scope enforcement, tool listing vs authorization, DCR scope changes, signed identity bridging, IDOR/allowlists, path normalization, token budget considerations, RST parsing separation, and repo autosync needs) and shifts the original sources section to Section 15.

FEATURE_SUGGESTIONS/FEATURE_MCP_CLAUDE_INTEGRATION.md

[qodo-code-review]

Code Review by Qodo

🐞 Bugs (0) 📘 Rule violations (0) 📎 Requirement gaps (0)

Great, no issues found!

Qodo reviewed your code and found no material issues that require review

_{To customize comments, go to the Qodo configuration screen, or learn more in the docs.}

[github-actions]

⏱️ Performance report

• Run: https://github.com/amirbiron/CodeBot/actions/runs/30683255433
• Artifact: perf-durations

(No performance test durations collected. Mark tests with @pytest.mark.performance.)

[github-actions]

📖 Documentation Preview

The documentation has been built successfully!

• 📦 Download Documentation Artifacts
• 🔍 Check the workflow summary for coverage report
• 📊 Build completed at: Sat, 01 Aug 2026 04:07:24 GMT

To view locally:

1. Download the artifacts
2. Extract the zip file
3. Open index.html in your browser

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!