Supervise an AI coding agent (Claude Code / Codex / ACP) overnight and keep control: block forbidden tool calls before they run, gate 'done' on real tests, checkpoint work to git, cap spend, and get a morning report where every line is sourced. Local-first, no telemetry. Rust daemon + Tauri app + Claude Code plugin.
Drift inferred · capture-to-capture
- HIGH code analysis flagged hidden prompt content in MalyStern/agentrelay
- HIGH code analysis flagged hidden prompt content in MalyStern/agentrelay
transport stdio counts 0 tools · 0 res
· 0 prompts
permission surface via code analysis
No tools enumerated yet for this server.
prompt-surface
shipped agent-instruction files + hidden-content / dangerous-code findings —
quoted from the analyzed source
analyzed commit 74437f3 · analyzer v33 · 1w ago
skills & prompt files 2
- hidden: readme: hidden-unicode MalyStern-agentrelay-74437f3/i18n/README.md :118
characters `‹U+2066›…‹U+2069›`. (+3 more lines)
evidence-backed
findings quoted directly from the published source artifact — not inferred
code files: 69
filesystem 4
- fs MalyStern-agentrelay-74437f3/i18n/check.mjs :18
import { readFileSync, readdirSync } from "node:fs"; - fs MalyStern-agentrelay-74437f3/plugin/bin/guard.mjs :18
import { readFileSync, accessSync } from 'node:fs' - fs MalyStern-agentrelay-74437f3/plugin/bin/session-check.mjs :10
import { accessSync } from 'node:fs' - fs MalyStern-agentrelay-74437f3/plugin/mcp/server.mjs :18
import { accessSync } from 'node:fs'
shell / exec 2
- shell MalyStern-agentrelay-74437f3/plugin/bin/guard.mjs :17
import { spawnSync } from 'node:child_process' - shell MalyStern-agentrelay-74437f3/plugin/mcp/server.mjs :17
import { spawnSync } from 'node:child_process'
declared dependencies 19
- @types/node@^22.10.2
- typescript@^5.7.2
- vitest@^3.2.4
- serde@1
- serde_json@1
- thiserror@2
- chrono@0.4
- regex@1
- once_cell@1
- rusqlite@0.32
- tokio@1
- async-trait@0.1
show 7 more
- futures@0.3
- async-stream@0.3
- keyring@3
- clap@4
- getrandom@0.2
- sha2@0.11.0
- ureq@3.3.0