Skip to content

docs(threat-model): claim the prompt-injection egress controls - #348

Merged
adeelahmad merged 1 commit into
mainfrom
claude/project-thread-mj7ggd
Oct 14, 2026
Merged

adeelahmad merged 1 commit into
mainfrom
claude/project-thread-mj7ggd

Conversation

@adeelahmad

@adeelahmad adeelahmad commented Oct 11, 2026 •

Copy link
Copy Markdown
Owner

Requested by Adeel · project thread

Description

Before: threat-model.md listed the Assistant's x-prompt-injection row as unresolved (Q26), with N-PROMPT disclaiming injection resistance and nothing about what an injection could carry out.

After:

Last of the prompt-injection series. Rebased on v0.8.1 and regenerated with the same script, so it sits beside the browser and mailbox rows that landed meanwhile.

Motivation and Context

AGENTS.md: new attack surface goes into threat-model.md. This series added controls, so the model now says which are claimed (egress) and which are only best-effort (marking, flagging), so a security report knows what counts.

Types of changes

  • Bug fix (non-breaking change which fixes an issue)
  • New feature (non-breaking change which adds functionality)
  • Breaking change (fix or feature that would cause existing functionality to change; ! in the title)
  • Documentation only.

Checklist:

  • The title is a Conventional Commit.
  • My code follows the code style of this project (make hooks checks it on commit).
  • I added or updated tests. (n/a: ci/validate_threat_model.py checks prose, YAML and JSON agree)
  • My change requires documentation updates, and I have updated docs/ accordingly. (the threat model itself)
  • Users need more than the title to understand it, so I wrote it under ## Unreleased in CHANGELOG.md. (n/a)

Done questions (AGENTS.md)

  • Goal: the threat model reflects the injection defenses; merged on green CI.
  • Agents: n/a, documentation.
  • Zero setup: n/a, documentation.
  • Fewest clicks and keystrokes: n/a, documentation.
  • Configurable: n/a, documentation (the settings it cites are in the web app).
  • Light: n/a, documentation.
  • Private and secure: threat model updated for the new controls and their limits.
  • Cost: n/a, documentation.
  • User decides: n/a, documentation.
  • Upgrades: n/a, no stored data.
  • Honest: claims only what is merged; marking and flagging stay unclaimed; each claim cites the docs.
  • Tests, docs and generated files: validated with ci/validate_threat_model.py; the YAML prose hash refreshed by its tooling.
  • Open standards: n/a.
  • Result reported with evidence; refine later: mark pipeline, template, workflow and graph-ask prompts; a classifier threshold setting; direct channel adapters call untrusted.defang.

🤖 Generated with Claude Code

https://claude.ai/code/session_01BVez82cREgCJtFk6yHReed

@adeelahmad adeelahmad self-assigned this Oct 11, 2026
P-INJECTION-EGRESS: once an answer has read text from outside Lens,
outbound tools and setup changes ask first, and chat-room answers keep
only the links people, the archive, tools or Lens wrote. Correctness-only,
a bypass is VALID-HARDENING (Q26). The Assistant x-prompt-injection row
is now claimed; N-PROMPT says marking and flagging make injection less
likely and are not claimed; MCP results and the optional classifier are
described; Q26 lists #302, #309, #329, #338, #343, #344 and #347.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01BVez82cREgCJtFk6yHReed
@adeelahmad
adeelahmad force-pushed the claude/project-thread-mj7ggd branch from 685d722 to 59ee829 Compare October 14, 2026 09:54
@adeelahmad
adeelahmad merged commit 80e102f into main Oct 14, 2026
11 checks passed
@adeelahmad
adeelahmad deleted the claude/project-thread-mj7ggd branch October 14, 2026 09:57
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants