docs(main): improve README discoverability for indirect prompt injection and MCP#21
Merged
Conversation
…d MCP - Update H1 to include keywords: indirect prompt injection, MCP tools - Update first paragraph to mention "indirect" and "MCP tool calls" - Add "protection" framing alongside "defense" Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
There was a problem hiding this comment.
Pull request overview
Updates the README heading and opening paragraph to improve SEO discoverability by adding keywords like "Indirect Prompt Injection", "MCP", and "Protection".
Changes:
- Updated H1 to include "Indirect Prompt Injection Protection for MCP Tools"
- Updated opening paragraph to mention "indirect", "MCP tool calls", and "protection"
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Updated project description for clarity and accuracy.
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
1 issue found across 1 file
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="README.md">
<violation number="1" location="README.md:1">
P2: Keep the package name casing consistent with the install/import examples; the mixed-case H1 currently documents a different identifier than `@stackone/defender`.</violation>
</file>
Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.
Contributor
There was a problem hiding this comment.
1 issue found across 1 file (changes from recent commits).
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="README.md">
<violation number="1" location="README.md:3">
P3: Change `tools calls` to `tool calls` in the opening sentence.</violation>
</file>
Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.
- Add StackOne logo, centered header, tagline - Add npm version, downloads, license, model size, latency, F1 badges - Add light/dark mode hero diagram using <picture> + prefers-color-scheme - Dark mode: SVG diagram from stackone.com; light mode: OG PNG Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Contributor
There was a problem hiding this comment.
1 issue found across 1 file (changes from recent commits).
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="README.md">
<violation number="1" location="README.md:22">
P3: The new image alt text makes a stricter latency claim (`under 2ms`) than the rest of the README supports. Use wording that matches the documented latency or remove the timing from the alt text.</violation>
</file>
Reply with feedback, questions, or to request a fix. Tag @cubic-dev-ai to re-run a review.
- Add assets/banner-light.svg and assets/banner-dark.svg: custom 1200×280px banners with full StackOne wordmark, "Defender" title, "@stackone/defender" subtitle, tagline, and stats pills (90.8% F1, 22MB, ~10ms, CPU-only) - Light banner uses brand surfaces (#FEFEFD bg, #047B43 accent/stats) - Dark banner uses dark surfaces (#2B2B2B bg, #76D6AE accent/stats) - Update README <picture> to reference new local assets instead of website images - Move badges below banner for cleaner visual hierarchy Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
glebedel
changed the title
- Add assets/demo-light.svg and assets/demo-dark.svg: 1200×400px diagrams showing the defender pipeline — poisoned email input → defender analysis → blocked DefenseResult with tier2Score: 0.97 and riskLevel: critical - Split badges into two rows: package info (npm, release, stars, license, TypeScript) + performance stats (model, latency, CPU-only, F1 score) - Add GitHub release badge (shows latest tagged release) - Add GitHub stars badge (social proof) - Add TypeScript typed badge Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Fix detections to use real pattern IDs: 'role_system_bracket' and 'ignore_previous' (matching src/classifiers/patterns.ts) - Fix fieldsSanitized to show correct array format: [ 'body' ] - Fix injection example to trigger actual patterns: [SYSTEM]: prefix triggers role_system_bracket; "Ignore all previous instructions" triggers ignore_previous - Fix shield proportions: 72×98px (1.36:1 ratio, up from 1.7:1) - Widen center panel (240px) for better shield breathing room - Improve code panel layout: consistent property/value alignment - Add maxSentence row showing actual matched sentence text - Remove fabricated 'command_injection' detection Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: cubic-dev-ai[bot] <191113872+cubic-dev-ai[bot]@users.noreply.github.com>
Shield was 72×108px (1.5:1, too tall). Now 92×92px (1.0:1). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
# @stackone/defenderto# @StackOne/defender — Indirect Prompt Injection Protection for MCP ToolsThese changes address gaps in search discoverability — the old README mentioned "prompt injection" but missed "indirect", had no "MCP" in the opening, and used "defense" but never "protection".
Also done (outside this PR, via API)
prompt-injection,indirect-prompt-injection,prompt-injection-defense,prompt-injection-protection,mcp-security,ai-security,llm-security,prompt-injection-detection🤖 Generated with Claude Code
Summary by cubic
Improves README SEO and clarity around indirect prompt injection and MCP with a branded banner, local light/dark diagrams, and a sharper intro focused on agents, MCP/CLI/direct tool calls, and protection for
@stackone/defender.New Features
<picture>with localassets/banner-*.svg; move badges below in two rows (npm/downloads/release/stars/license/TypeScript + model/latency/CPU-only/F1).Bug Fixes
role_system_bracketandignore_previous,fieldsSanitized: [ 'body' ], addmaxSentence, remove non-existent rule.Written for commit 9337445. Summary will update on new commits.