Scripthold — Secure MCP Server for Local Workspaces

Code from the web. Work locally. Recover safely.

Scripthold is a Model Context Protocol (MCP) server that gives web, desktop, and CLI agents controlled access to explicitly authorized local workspaces. It safely handles legacy text encodings, deterministic file changes, source navigation, backups, and optional durable local execution over stdio or authenticated Streamable HTTP.

AI clients see Настройки — not ???? or Íàñòðîéêè.

What Scripthold provides

The current public release is Scripthold 3.3.0 with 43 tools and 3 guided prompts.

• 168 registered text encodings with content-based detection, UTF-32 LE/BE support, and conservative ambiguity handling.

• 100 active Source Intelligence providers for bounded declaration navigation, structural search, selected project relations, and verified context assembly.

• Secure workspace boundaries with canonical-root containment, Windows reparse/junction handling, deterministic traversal, and missing-path validation.

• Approval-bound mutations using preview/apply capabilities, exact fingerprints, conflict checks, staged writes, and truthful partial-state reporting.

• Optional persistent backups with history, compare, audit, restore, explicit garbage collection, pinning, and exact-ID deletion.

• Durable asynchronous tasks with idempotency, persistent state, bounded logs, recovery, locks, and cancellation.

• Reliable long-running read-only calls through optional deferred-operation storage and bounded retained-result retrieval.

• Authenticated Streamable HTTP with loopback defaults, bearer authentication, Host/Origin validation, resource limits, and explicit TLS/proxy requirements for non-loopback exposure.

Scripthold originated from the original mcp-file-tools project, created by Dimitar Grigorov, and retains its GPL-3.0 lineage. See Project Direction.

Related MCP server: codex-web-bridge

Quick start

Use a published release

Download the asset for your platform from the latest release and verify it against checksums.txt. Releases also attach a signed Sigstore build-provenance bundle for the normal GoReleaser assets.

Build from source

git clone https://github.com/zoster81/scripthold.git
cd scripthold
go test ./...
go build -o scripthold ./cmd/scripthold

The module path is github.com/zoster81/scripthold.

Local stdio client

Pass each authorized directory as a startup argument:

{
  "mcpServers": {
    "scripthold": {
      "type": "stdio",
      "command": "C:\\Tools\\scripthold_windows_amd64.exe",
      "args": ["D:\\Projects"]
    }
  }
}

Startup roots are authoritative. A roots-capable stdio client may supply dynamic roots only when the process starts without directory arguments.

Verified self-update

A supported standalone binary can opt in to the verified self-update lifecycle with scripthold update --adopt. Before running it, stop every other Scripthold process using that binary and keep them stopped until the adoption command exits. Adoption verifies the current standalone installation and creates its local stable self-update state; it performs no network work, downloads no release, and does not replace the executable. MCPB-managed, linked/aliased, unsupported, dirty, or already-adopted installations fail closed.

After adoption, run scripthold update to check the latest immutable release, verify the matching platform binary and release evidence, stage a candidate, and start the detached update helper. If no newer release exists, the command succeeds without changing the installation. A successful dispatch means the helper has started; use scripthold update status to observe completion before starting another update. If the command reports that the helper started but launcher cleanup failed, inspect status before retrying.

scripthold update status is read-only and does not start the MCP server. It reports adoption, observed update state, durable/current/candidate version evidence, release availability, and whether explicit recovery is available. If the installed binary cannot be assigned a verified semantic version, the remote comparison is suppressed rather than using stale evidence. MCP_NO_UPDATE_CHECK=1 disables this status-time remote check, in which case the latest version is reported as unknown.

When status reports that recovery is available, run scripthold update recover to dispatch the fixed verified recovery helper. Recovery performs no network work and derives authority from the durable transaction plus observed installation bytes. A successful dispatch means the recovery helper has started; use status to observe completion, and if launcher cleanup fails after dispatch, inspect status before retrying.

The update word is reserved as a command namespace; use -- update if a literal authorized directory is named update.

Streamable HTTP

HTTP requires exactly one bearer-token source. A minimal loopback PowerShell launch is:

$tokenPath = Join-Path $env:TEMP "scripthold.token"
$bytes = New-Object byte[] 32
$rng = [System.Security.Cryptography.RandomNumberGenerator]::Create()
try { $rng.GetBytes($bytes) } finally { $rng.Dispose() }
[System.IO.File]::WriteAllText($tokenPath, [Convert]::ToBase64String($bytes), [System.Text.UTF8Encoding]::new($false))

$env:MCP_HTTP_TOKEN_FILE = $tokenPath
$env:MCP_HTTP_ADDR = "127.0.0.1:8765"
.\scripthold_windows_amd64.exe --transport=streamable-http D:\Projects

The MCP endpoint is http://127.0.0.1:8765/mcp; /healthz and /readyz expose liveness/readiness. Send the token as Authorization: Bearer <token> on every MCP request.

Do not expose HTTP beyond loopback without reading HTTP Security. Non-loopback use requires explicit opt-in and TLS or a trusted proxy boundary.

OpenAI Secure MCP Tunnel

Sanitized PowerShell examples are available under examples/:

• start-local-stdio.ps1

• start-local-http.ps1

• start-openai-tunnel-stdio-plus-local-http.ps1

• start-openai-tunnel-http-plus-local-stdio.ps1

Copy an example outside the checkout before replacing placeholders. Never commit Runtime API keys, Tunnel IDs, bearer tokens, or private state paths.

Container

docker build --build-arg VERSION=dev -t scripthold:dev .

docker run --rm -i \
  --read-only \
  --cap-drop=ALL \
  --security-opt=no-new-privileges \
  --tmpfs /tmp:rw,noexec,nosuid,size=64m \
  --mount type=bind,source=/absolute/project,target=/data \
  scripthold:dev --transport=stdio /data

The image runs as unprivileged UID/GID 10001.

Tool groups

TOOLS.md is the detailed public reference for schemas, parameters, examples, limits, and error behavior. The table below reflects the Scripthold 3.3.0 public catalog.

Source Intelligence

Source Intelligence is read-only. It does not execute project code and does not require external parser/compiler/LSP processes.

source_symbols provides bounded outline, digest, find, and fingerprint-bound show. source_query provides structural search, supported project relations, dependency graphs, and fingerprint-verified context. Unsupported or ambiguous relationships fail closed rather than being guessed.

Markdown is intentionally outside Source Intelligence. The dedicated Markdown tools understand document structure instead of treating a Markdown file as source code.

Markdown and documentation workflows

Use markdown_read when an agent needs to understand a Markdown document before changing it. It can inspect headings and sections, find structural elements, resolve local fragments and links, validate managed TOCs, and return exact authored source for a selected structure.

Use markdown_workspace when the question spans several Markdown files. It can scan an authorized documentation tree or follow local Markdown relationships from explicit entry files, then inspect document inventory, query graph edges/backlinks/reachability, validate workspace problems, and report Marksplice-proven repair plans for explicitly managed TOCs. It is read-only and never repairs files directly.

Use markdown_create to prepare a complete new Markdown document at an authorized path that does not yet exist. It builds typed front matter, blocks, inline content, lists, references, footnotes, math, and tables through Marksplice, returns the exact generated Markdown and physical-file metadata for review, and does not create the file until its one-shot preview is passed to markdown_apply.

Use markdown_edit to prepare source-preserving changes without writing the file. It covers document structure such as headings, sections, paragraphs, list items, tasks, table alignment, and managed TOCs; inline and reference edits; front matter; fenced code; blockquotes and alerts; footnotes; math; raw HTML; and reviewed existing-document constructors. A preview can contain up to 64 compatible operations against one immutable snapshot. Marksplice validates the combined final state: local fragments must remain on the same semantic target or be explicitly retargeted, and a sync/toc operation can be combined with other edits so the TOC is regenerated from the final candidate. Conflicting, overlapping, or unsafe changes fail closed. See Tool Reference for the exact closed operation forms.

Use markdown_apply only after reviewing a markdown_edit or markdown_create preview. Its only input is the returned previewId. Existing-file edits revalidate the approved source and backup policy; new-file creation revalidates the missing destination and its parent identity and uses no-replace creation. Both paths report the actual final state. A typical workflow is therefore: read the structure → prepare a preview → review it → apply it.

This is useful for tasks such as correcting section names, promoting or demoting headings, adding guidance next to an existing paragraph, rewriting or removing obsolete prose, deleting an obsolete section together with its subsections, validating local anchors before editing, or updating documentation while preserving the file's existing formatting and line endings. Markdown semantic text is canonical UTF-8; the mutation bridge also preserves UTF-16 LE/BE and UTF-32 LE/BE compatibility files after an exact source round-trip check, while other unproven encodings fail safely instead of being silently rewritten.

The generated Source Intelligence provider/capability matrix is in Language Capabilities. Detailed Markdown schemas and examples are in TOOLS.md.

Safety model

• File access is limited to explicitly authorized roots after canonical path validation.

• Symlink, junction, reparse-point, alias, and missing-path escapes fail closed.

• Encoding detection uses bytes and decoded-content evidence, never filenames or extensions.

• Mutations stage and revalidate before commit; initially missing destinations use no-replace semantics.

• Preview/apply workflows bind the approved operation to a one-shot capability so mutation parameters cannot be changed at apply time.

• Multi-file mutations do not claim transactional rollback; failures report committed/unchanged/unknown state where applicable.

• The optional backup store is a separate protected authority outside public roots.

• task_run execution is disabled by default. Shell and script execution require explicit authorization; HTTP requires an additional HTTP execution opt-in.

• HTTP is authenticated and loopback-only by default.

Current architectural boundaries are summarized in Architecture.

Key configuration

Most installations need only a small subset of environment variables:

Detailed limits and subsystem-specific options are documented where they are used: TOOLS.md, Durable Tasks, and HTTP Security.

Typical uses

• Safely edit legacy source/configuration files without corrupting encoding or line endings.

• Search mixed-encoding repositories with explicit partial-coverage evidence.

• Navigate heterogeneous source trees without loading entire projects into the model.

• Preview and approve exact single- or multi-file changes.

• Keep persistent pre-state backups and restore selected versions.

• Run long builds/tests without tying process lifetime to one MCP request.

• Serve the same workspace toolset through local stdio, authenticated HTTP, containers, or a secure tunnel bridge.

Documentation

• Tool Reference — public MCP tools, schemas, examples, and limits.

• Markdown Intelligence — Markdown semantics, safety boundaries, and Marksplice capability mapping.

• Architecture — current product and security boundaries.

• HTTP Security — HTTP deployment and threat model.

• Durable Tasks — persistent task execution.

• Language Capabilities — generated Source Intelligence matrix.

• Roadmap — current and future work only.

• Migration 2.0 / Migration 3.0 — intentional compatibility changes.

• Contributing / Publishing / Security — maintainer and contributor guidance.

• Changelog — concise release and unreleased user-visible changes.

Development

The required Go version is declared by go.mod.

go mod verify
go test ./...
golangci-lint run ./...
go build -o scripthold ./cmd/scripthold

See CONTRIBUTING.md before proposing changes.

License

GPL-3.0 — see LICENSE.