Exteragram MCP Server
@catalystdev/exteragram-mcp
npm
v1.0.0
Published by @catalystdev — no publish provenance, so origin is unverified, but the source is public: the repository link below is self-declared yet readable, so you can inspect the code before adopting it.
The grade answers one question — how safe is this server for you to adopt — so it is computed in two auditable stages. Nothing below is an opinion or an LLM's guess; every line is a real term the deterministic engine applied, and the same input always yields the same number.
1. Threat score — 100 − 54.7 = 50. What the published surface and source actually contain:
| Points | What was found | Category |
|---|---|---|
| −54.7 | Mixed-script (homoglyph) text in description ×21 MTC-UNI-009 | injection |
2. Client adoption risk — 50 − 7 = 43. Three small, subtract-only factors that reflect your risk in adopting it — a clean scan proves less on a powerful, unverified or barely-inspectable package, so the grade says so plainly:
| Points | Adoption-risk factor |
|---|---|
| −6 | capability blast radius (high) — client exposure if the model is manipulated |
| −1 | publisher verification (public source) — no provenance, but the source is public and inspectable |
| ✓ | inspection depth (source) — how much of the target the scan could see |
Capability observations and info notes are shown under Findings but never scored.
Open any row's finding below for the file, line and evidence behind a deduction.
Findings 46
high Sensitive-source and external-sink co-existMTC-FLOW-004
Tools that read sensitive data ([adb_deploy_plugin, adb_get_logs, adb_push_file, adb_pull_file, generate_edit_message, set_plugin_context, generate_read_file, generate_write_file, generate_delete_file, generate_ensure_dir, generate_list_dir]) and tools that can send data out ([adb_remove_plugin, adb_shell, generate_send_message, generate_send_request, generate_send_message_hook, generate_pre_request_hook, generate_post_request_hook]) are exposed together. An agent can move private data to the sink.
Fix: Keep secret-reading and egress capabilities on separate, separately-approved servers.
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Выполняет 'adb devices' и возвращает список подключённых Android-устройств.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool adb_check_devices · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует R(callable) — обёртку Python-функции в Java Runnable.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_runnable · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует Python-класс с @java_subclass декоратором, создающий реальный Java-подкласс.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_java_subclass · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует метод с @joverride или @joverload для переопределения Java-метода.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_joverride · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует jfield(...) — декларацию Java-поля с опциональными getter/setter.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_jfield · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует jMVELmethod или jMVELoverride — Java-метод с MVEL-кодом вместо Python.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_jmvel_method · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует доступ к Java-объекту через J() (JavaHelper).
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_java_helper · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует new_instance() или new_java_instance() для создания объекта Java-подкласса.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_new_instance · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует PyObj.create(python_object) — оборачивает Python-объект
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_pyobj · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует вызов send_request(request, callback) для произвольного Telegram TL-запроса.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_send_request · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Возвращает полный список getter-функций для доступа к контроллерам Telegram.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool list_controllers · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Возвращает все стандартные getter-функции SDK с реальными путями на устройстве.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool list_standard_dirs · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Возвращает список pre-installed Python-библиотек и инструкцию
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool list_available_libraries · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Возвращает FQN-пути наиболее используемых Java-классов Telegram и Android.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool list_common_classes · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Возвращает готовый Python-код шаблона плагина.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool get_plugin_template · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует хук перехвата Telegram-запроса ДО отправки на сервер.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_pre_request_hook · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует хук обработки ответа Telegram-запроса ПОСЛЕ получения ответа.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_post_request_hook · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует хук перехвата отдельного Telegram-обновления.
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_on_update_hook · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует хук перехвата контейнера Telegram-обновлений (Updates).
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_on_updates_hook · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Возвращает таблицу всех поддерживаемых HTML-тегов и Markdown-синтаксиса
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool list_html_tags · description
high Mixed-script (homoglyph) text in descriptionMTC-UNI-009
A single token mixes multiple scripts (Latin, Cyrillic). This is the signature of a homoglyph attack — visually identical characters from another script used to impersonate a trusted name or slip past keyword filters.
Evidence: Генерирует полный паттерн Xposed-хука в 3 шага:
Fix: Restrict metadata to a single script, or normalize confusables before display.
Location: tool generate_method_hook · description
high Tool "adb_remove_plugin" exposes command/code executionMTC-CAP-001
Tool "adb_remove_plugin" appears to run shell commands or evaluate code (keyword "shell"). Arbitrary execution driven by model input is one of the most dangerous MCP capabilities; combined with any untrusted input it becomes RCE.
Fix: Sandbox execution, allowlist commands/arguments, and never pass model output to a shell unescaped.
Location: tool adb_remove_plugin
high Tool "adb_shell" exposes command/code executionMTC-CAP-001
Tool "adb_shell" appears to run shell commands or evaluate code (keyword "shell", parameter "command"). Arbitrary execution driven by model input is one of the most dangerous MCP capabilities; combined with any untrusted input it becomes RCE.
Fix: Sandbox execution, allowlist commands/arguments, and never pass model output to a shell unescaped.
Location: tool adb_shell
high Shell/command execution in server code (dist/tools/adb.d.ts)MTC-SRC-002
In the server's implementation (`dist/tools/adb.d.ts:11`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: child_process
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server dist/tools/adb.d.ts
high Shell/command execution in server code (dist/tools/adb.js)MTC-SRC-002
In the server's implementation (`dist/tools/adb.js:11`): Spawning a shell/process is command-execution capability; with unsanitized tool input it is command injection / RCE. This is read from the code itself — not from the tool description — so a poisoned server cannot hide it behind honest-looking metadata.
Evidence: child_process
Fix: Review this call path: confirm it never receives unsanitized tool input, constrain it, or remove it. Treat a server whose code reaches these sinks as high-capability regardless of what its tools claim.
Location: server dist/tools/adb.js
medium Tool "generate_write_file" can modify the filesystemMTC-CAP-002
Tool "generate_write_file" can write, overwrite or delete files (keyword "write_file"). Verify it is scoped to a safe directory.
Fix: Constrain file operations to an explicit, non-sensitive root; reject path traversal.
Location: tool generate_write_file
medium Tool "generate_delete_file" can modify the filesystemMTC-CAP-002
Tool "generate_delete_file" can write, overwrite or delete files (keyword "delete_file"). Verify it is scoped to a safe directory.
Fix: Constrain file operations to an explicit, non-sensitive root; reject path traversal.
Location: tool generate_delete_file
medium Unconstrained command parameter "command" on "adb_shell"MTC-CAP-006
Tool "adb_shell" takes a command-shaped parameter "command" with no enum/pattern constraint. Free-form, model- or attacker-controlled arguments reaching a shell is the command-injection precondition.
Fix: Constrain the parameter (enum/pattern), or build the command from a fixed template with escaped args.
Location: tool adb_shell · inputSchema.properties.command
low Mutating tool "adb_remove_plugin" declares no destructiveHintMTC-CAP-005
Tool "adb_remove_plugin" can mutate/egress but declares no destructiveHint. Clients that don't default to spec-safe behavior may not prompt before running it.
Fix: Declare accurate annotations, and gate destructive tools on user confirmation regardless.
Location: tool adb_remove_plugin
low Mutating tool "adb_shell" declares no destructiveHintMTC-CAP-005
Tool "adb_shell" can mutate/egress but declares no destructiveHint. Clients that don't default to spec-safe behavior may not prompt before running it.
Fix: Declare accurate annotations, and gate destructive tools on user confirmation regardless.
Location: tool adb_shell
low Mutating tool "generate_write_file" declares no destructiveHintMTC-CAP-005
Tool "generate_write_file" can mutate/egress but declares no destructiveHint. Clients that don't default to spec-safe behavior may not prompt before running it.
Fix: Declare accurate annotations, and gate destructive tools on user confirmation regardless.
Location: tool generate_write_file
low Mutating tool "generate_delete_file" declares no destructiveHintMTC-CAP-005
Tool "generate_delete_file" can mutate/egress but declares no destructiveHint. Clients that don't default to spec-safe behavior may not prompt before running it.
Fix: Declare accurate annotations, and gate destructive tools on user confirmation regardless.
Location: tool generate_delete_file
low Unconstrained path parameter "file_path" on "adb_deploy_plugin"MTC-CAP-008
Tool "adb_deploy_plugin" takes a path parameter "file_path" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool adb_deploy_plugin · inputSchema.properties.file_path
low Unconstrained path parameter "local_path" on "adb_push_file"MTC-CAP-008
Tool "adb_push_file" takes a path parameter "local_path" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool adb_push_file · inputSchema.properties.local_path
low Unconstrained path parameter "remote_path" on "adb_push_file"MTC-CAP-008
Tool "adb_push_file" takes a path parameter "remote_path" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool adb_push_file · inputSchema.properties.remote_path
low Unconstrained path parameter "remote_path" on "adb_pull_file"MTC-CAP-008
Tool "adb_pull_file" takes a path parameter "remote_path" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool adb_pull_file · inputSchema.properties.remote_path
low Unconstrained path parameter "local_path" on "adb_pull_file"MTC-CAP-008
Tool "adb_pull_file" takes a path parameter "local_path" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool adb_pull_file · inputSchema.properties.local_path
low Unconstrained path parameter "file_path" on "generate_edit_message"MTC-CAP-008
Tool "generate_edit_message" takes a path parameter "file_path" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool generate_edit_message · inputSchema.properties.file_path
low Unconstrained path parameter "file_path" on "set_plugin_context"MTC-CAP-008
Tool "set_plugin_context" takes a path parameter "file_path" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool set_plugin_context · inputSchema.properties.file_path
low Unconstrained path parameter "path_expr" on "generate_read_file"MTC-CAP-008
Tool "generate_read_file" takes a path parameter "path_expr" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool generate_read_file · inputSchema.properties.path_expr
low Unconstrained path parameter "path_expr" on "generate_write_file"MTC-CAP-008
Tool "generate_write_file" takes a path parameter "path_expr" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool generate_write_file · inputSchema.properties.path_expr
low Unconstrained path parameter "ensure_dir" on "generate_write_file"MTC-CAP-008
Tool "generate_write_file" takes a path parameter "ensure_dir" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool generate_write_file · inputSchema.properties.ensure_dir
low Unconstrained path parameter "path_expr" on "generate_delete_file"MTC-CAP-008
Tool "generate_delete_file" takes a path parameter "path_expr" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool generate_delete_file · inputSchema.properties.path_expr
low Unconstrained path parameter "path_expr" on "generate_ensure_dir"MTC-CAP-008
Tool "generate_ensure_dir" takes a path parameter "path_expr" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool generate_ensure_dir · inputSchema.properties.path_expr
low Unconstrained path parameter "path_expr" on "generate_list_dir"MTC-CAP-008
Tool "generate_list_dir" takes a path parameter "path_expr" with no constraint. Without a canonicalize-and-contain check (not visible statically), this permits ../ traversal outside the intended root.
Fix: Resolve and verify the path stays within an allowed root; reject traversal sequences.
Location: tool generate_list_dir · inputSchema.properties.path_expr
Toxic flows 1
Cross-tool combinations that form a data-exfiltration primitive (untrusted input → sensitive source → external sink).
What this scan could not see
- Tools were statically extracted from the published source (81 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".
Versions
Scan history per published version. The engine is deterministic — the same version always yields the same score, so a changed score means the package itself changed.
| Version | Score | Findings | Engine | Scanned |
|---|---|---|---|---|
v1.0.0 latest |
F 43/100 | 46 | 1.7.0 | 2026-07-23 |
Embed this score
Show this server's live Trust Score in your README, docs or website. The badge is served straight from the registry and updates automatically after every rescan — no API key needed. It links back to this page, so anyone who sees the grade can also read the findings behind it instead of taking a number on faith.
Verify this score yourself
The score above is reproducible: the same package version always yields the same result. Run it locally or over the free API — no account, no LLM, fully deterministic.
npx mcptrustchecker scan @catalystdev/exteragram-mcp --onlineMore in Developer Tools
Security scan results for the 1stay MCP server.
Security scan results for the Adguard Home MCP server.
Security scan results for the Ado Browser MCP server.
Security scan results for the Aemet MCP server.
Security scan results for the Agent MCP server.
Security scan results for the Agent Recorder MCP server.