TOOLS.md

Native Tool Inventory

Generated from the MaskShift 1.4.1 runtime. 202 native tools are available before plugins or MCP servers add more capabilities.

Only activated descriptors enter a model request; this document is the complete local catalog.

agent-bridge (4)

Tool	Access	Risk	Description
agent_bridge_discover	read	normal	Detect locally installed Claude Code, Codex, OpenCode, Copilot, Hermes, Aider, and configured coding-agent CLIs.
agent_bridge_help	read	normal	Read the installed command help for an external coding-agent bridge before delegating.
agent_bridge_run	write	host-exec	Run an installed Claude Code, Codex, OpenCode, Copilot, Hermes, Aider, or configured agent against the current workspace. Can wait or return a persistent process.
external_agent_run	write	host-exec	Execute any configured or ad-hoc coding-agent command with placeholder arguments such as {prompt}, {cwd}, and {workspace}.

artifacts (5)

Tool	Access	Risk	Description
archive_create	write	write	Create tar.gz, tar.zst, tar, or zip archives from arbitrary host paths.
archive_extract	write	write	Extract zip, tar.gz, tar.zst, tar, and common compressed archives.
archive_list	read	normal	List the files inside a zip or tar archive without extracting it.
checksum_verify	read	normal	Compute a file's hash and compare it against an expected digest, returning a match/mismatch verdict instead of leaving the comparison to the model.
file_hash	read	normal	Calculate SHA-256, SHA-512, SHA-1, or MD5 for a file without loading it all into model context.

automation (7)

Tool	Access	Risk	Description
automation_create	write	persistent-exec	Schedule a recurring or one-time autonomous agent prompt, tool call, or unrestricted shell command. Schedules accept cron, ISO timestamps, or strings like every 15m.
automation_delete	write	write	Permanently remove a scheduled automation.
automation_list	read	normal	List scheduled agent, tool, and shell automations with next/last run state.
automation_pause	write	persistent-exec	Disable an automation without deleting it.
automation_resume	write	persistent-exec	Enable an automation and compute its next run.
automation_run_now	write	host-exec	Immediately execute an automation regardless of its next scheduled time.
automation_update	write	persistent-exec	Edit an automation schedule, action, name, metadata, or enabled state.

browser (23)

Tool	Access	Risk	Description
browser_accessibility	read	normal	Return the Chrome accessibility tree for semantic page understanding.
browser_act	write	external-action	Click, fill, select, check or press a key on an element by its ref from browser_extract. Waits for the page to settle and returns the new URL. Refuses to click purchase-style buttons ("Place order", "Pay now") unless the person confirms; refuses to type into password fields (use browser_login).
browser_choose	write	external-action	Extract the options on the current page, show them to the person as a picker and return what they chose. With open:true the chosen option is opened. Use this when a decision is theirs: which restaurant, product, flight or result.
browser_click	write	external-action	Click a CSS selector or page coordinate through the Chrome input pipeline.
browser_close	write	process	Terminate a MaskShift browser instance and its CDP connections.
browser_close_tab	write	process	Close one browser page target.
browser_console	read	normal	Read recent console messages, exceptions, and browser log entries from the page.
browser_discover	read	normal	Detect an installed Chromium, Chrome, or Edge executable for dependency-free CDP automation.
browser_evaluate	write	external-action	Execute arbitrary JavaScript in the active page and return a serializable value.
browser_extract	read	normal	Describe the current page as selectable options (restaurants, products, results), forms, blockers (login wall, CAPTCHA, cookie notice) and actions, each with a short ref such as e12 for browser_act. Far smaller than the raw page; prefer it to browser_snapshot for web tasks.
browser_handoff	write	external-action	Ask the person to do something in the live browser view that you should not (solve a CAPTCHA, approve a bank prompt, enter payment details) and wait until they say they are done.
browser_instances	read	normal	List browser processes launched by MaskShift.
browser_launch	write	process	Launch a persistent-profile Chromium browser with DevTools automation. Headless by default; visible mode supports interactive logins.
browser_login	write	external-action	Sign in to the site open in the browser. The person is asked for their username, password and any verification code in the terminal; the values are typed into the page directly and never shown to you. CAPTCHAs are handed to the person in the live browser view. Returns only the outcome.
browser_navigate	write	network	Navigate a browser tab and wait for the document to load.
browser_network	read	normal	Read recent Chrome DevTools Network events for the page.
browser_new_tab	write	network	Create a new browser tab at a URL.
browser_print_pdf	read	normal	Render the current browser page to a PDF artifact using Chrome print layout.
browser_screenshot	read	normal	Capture a viewport or full-page screenshot to a workspace artifact file.
browser_snapshot	read	normal	Return page title, URL, visible text, and a selector map of interactive elements.
browser_tabs	read	normal	List all page targets in a browser instance.
browser_type	write	external-action	Focus an optional CSS selector, type text, and optionally submit with Enter.
browser_wait_for	read	normal	Wait for a CSS selector to become visible, hidden, attached, or detached.

code-intelligence (12)

Tool	Access	Risk	Description
lsp_code_actions	write	write	List the refactors and quick fixes the language server offers at a position or range (fix this error, extract a function, add a missing import…) and optionally apply one by index, by title, or the preferred one. Prefer this over hand-editing for mechanical refactors.
lsp_definition	read	normal	Resolve the definition location for a symbol at a source position.
lsp_diagnostics	read	normal	Return language-server errors, warnings, hints, and related information for a file.
lsp_discover	read	normal	Detect installed language servers for TypeScript, Python, C/C++, Rust, Go, Java, Ruby, Lua, JSON, HTML, CSS, and YAML.
lsp_format	write	write	Ask the language server to format a document and optionally apply the edits.
lsp_hover	read	normal	Get type, signature, and documentation information at a 1-based source position.
lsp_organize_imports	write	write	Sort and prune a file's imports with the language server.
lsp_references	read	normal	Find language-aware references for a symbol at a source position.
lsp_rename	write	write	Compute and optionally apply a workspace-wide language-aware symbol rename.
lsp_status	read	normal	List active workspace language server processes and capabilities.
lsp_stop	write	process	Stop one or all lazy language server processes for the workspace.
lsp_symbols	read	normal	Return language-server document symbols and hierarchy for a source file.

containers (9)

Tool	Access	Risk	Description
container_build	write	host-exec	Build a Docker/Podman image from a Dockerfile or Containerfile with build args and tags.
container_compose	write	host-exec	Run Docker Compose or Podman Compose actions such as up, down, build, logs, ps, and config.
container_engine	read	normal	Detect Docker or Podman and return version and runtime information.
container_exec	write	host-exec	Execute a command in an existing Docker/Podman container.
container_list	read	normal	List running or stopped Docker/Podman containers as structured records.
container_logs	read	normal	Read recent container logs with timestamps, tail, and since controls.
container_run	write	host-exec	Run any Docker/Podman image with ports, volumes, environment, network, privilege, and detach controls.
container_stop	write	host-exec	Stop, kill, restart, pause, unpause, or remove a container.
kubernetes_exec	write	remote-exec	Execute any kubectl operation with a structured argument list and optional context, namespace, and kubeconfig.

database (3)

Tool	Access	Risk	Description
database_cli	write	database-write	Execute a command through psql, mysql, redis-cli, mongosh, duckdb, or another installed database client.
sqlite_query	write	database-write	Open any SQLite database directly through Node native SQLite, execute parameterized SQL, and return structured rows. Write statements are allowed.
sqlite_schema	read	normal	Return tables, views, indexes, triggers, and CREATE statements from a SQLite database.

documents (5)

Tool	Access	Risk	Description
doc_read	read	normal	Read the text of a document with no external tools: PDF, Word (.docx), Excel (.xlsx, sheets rendered as tables), PowerPoint (.pptx, with speaker notes), OpenDocument (.odt/.ods/.odp), EPUB, RTF and HTML. Use this instead of fs_read for any non-plain-text document. Legacy binary .doc/.xls/.ppt must be re-saved in the modern format first.
image_read	read	normal	Read an image file: extract any visible text with OCR, and — if a local vision model is available — generate a natural-language description or answer a question about it. Works even when the active chat model has no native vision support, because the description is produced out of band and returned as plain text.
notebook_edit	write	write	Replace, insert, or delete one cell in a Jupyter (.ipynb) notebook by index. Replacing or inserting a code cell clears its stale outputs and execution count.
notebook_read	read	normal	Read a Jupyter (.ipynb) notebook and return each cell's index, type, source, and a bounded summary of its outputs.
pdf_read	read	normal	Extract text from a PDF with a built-in parser (no external tools needed), with an optional page range. If the PDF is a scan with no text layer, falls back to rendering pages and running OCR when poppler-utils and tesseract are installed; also falls back to pdftotext if the built-in parser cannot read the file.

filesystem (16)

Tool	Access	Risk	Description
chmod_set	write	write	Change a path's POSIX mode (chmod) and optionally its owner/group (chown), reporting the mode before and after.
file_diff	read	normal	Show a unified diff between two arbitrary paths (files or directories), independent of Git — for comparing anything on disk, not just a repo working tree.
fs_apply_patch	write	write	Apply a unified diff using git apply. The patch can update multiple files and is checked before application.
fs_delete	write	destructive	Delete a file or directory recursively. Autonomous mode executes immediately without an approval prompt.
fs_list	read	normal	List a directory tree with file sizes and types. Paths may be workspace-relative or absolute in autonomous mode.
fs_mkdir	write	write	Create a directory and missing parent directories.
fs_move	write	write	Move or rename a file or directory, optionally replacing the destination.
fs_patch	write	write	Apply one or more oldText/newText replacements to a file atomically. Minor whitespace or indentation differences are tolerated; text that is missing or ambiguous fails with the closest region shown, unless replaceAll is requested.
fs_read	read	normal	Read a UTF-8 text file with optional 1-based line range and hard output bounds.
fs_read_binary	read	normal	Read a bounded binary file as base64 with MIME-relevant metadata.
fs_replace_lines	write	write	Replace lines startLine..endLine (1-based, inclusive, as shown by fs_read) with new text. Pass expect (the first old line) to refuse the edit if the file has moved. Use an empty newText to delete lines, or endLine = startLine - 1 style ranges via fs_patch to insert.
fs_stat	read	normal	Return file type, size, timestamps, mode, target and hash metadata.
fs_write	write	write	Create or overwrite a file atomically. Parent directories are created automatically.
symbol_read	read	normal	Return one definition (function, class, method as Class.method) with its line range, without reading the whole file. With no symbol, list the definitions in the file.
symbol_replace	write	write	Replace one definition by name with new source. You supply only the new text; the old text is located for you, so it never has to be reproduced exactly. Use Class.method for methods.
text_stats	read	normal	Report line, word, and byte counts for one or more text files (like wc) without spending context on their contents — use before deciding whether to read a file in full.

fleet (9)

Tool	Access	Risk	Description
fleet_ask	write	agent	Give one member a task or question and wait for its reply. Its inbox is delivered with it. Any [[send]] messages it writes are queued for the teammates it addresses (use fleet_relay to have them acted on).
fleet_harnesses	read	normal	List the agent harnesses a fleet member can run on (claude, codex, opencode, hermes, copilot, aider, any configured custom bridge, or maskshift (the built-in engine)) and whether each is installed.
fleet_list	read	normal	List the fleet members with their harness, role, status, unread mail and last reply. Pass name for one member's full history.
fleet_messages	read	normal	Read the recent messages passed between fleet members (optionally just those to or from one member).
fleet_relay	write	agent	Hand a task to the fleet and let its members work it out together: the lead gets the task, members message each other with [[send]] blocks, and rounds continue until someone writes [[done]], everyone falls quiet, or the round/time limit is reached. Returns the outcome and a per-turn trace.
fleet_send	write	state	Queue a message from one member (or you) to another, or to="*" for everyone, without running anyone yet. It is delivered the next time the recipient takes a turn.
fleet_spawn	write	agent	Create a persistent, named fleet member backed by claude, codex, opencode, hermes, copilot, aider, any configured custom bridge, or maskshift (the built-in engine). Several members may share a harness. Members keep their history and inbox between turns and can message each other. Use isolated=true to give an editing member its own Git worktree.
fleet_stop	write	agent	Cancel a member's current turn, optionally removing it from the fleet. With no name, cancels every member and running relay.
fleet_suggest	read	normal	Rank the installed agent harnesses (Claude Code, Codex, OpenCode, Hermes, …) for a task by how each has actually done on similar tasks before, with the evidence behind the ranking.

git (10)

Tool	Access	Risk	Description
git_branch	write	write	List, create, switch, rename, or delete branches. MaskShift runs these directly without permission prompts.
git_checkpoint_create	write	local-snapshot	Snapshot tracked and untracked workspace changes so an agent run can be rolled back without interrupting normal Git history.
git_checkpoint_list	read	normal	List reversible MaskShift workspace checkpoints.
git_checkpoint_restore	write	destructive	Restore a prior MaskShift checkpoint. This is destructive to current workspace changes.
git_commit	write	write	Stage selected or all changes and create a commit. Optional amend and no-verify modes are supported.
git_diff	read	normal	Show working-tree, staged, commit-range, or selected-file diffs with configurable context and output bounds.
git_log	read	normal	Read compact commit history, optionally for a branch, range, author, grep expression, or path.
git_show	read	normal	Show a commit, tag, tree, or file at a revision.
git_status	read	normal	Inspect branch, upstream, staged, modified, deleted, renamed, conflicted, and untracked files.
git_worktree_create	write	write	Create a branch-backed Git worktree for isolated subagent or experimental work.

interaction (5)

Tool	Access	Risk	Description
credentials_forget	write	write	Remove the remembered sign-in for a site (from memory and the keychain).
credentials_list	read	normal	List which sites have a sign-in remembered, and where (this session or the system keychain). Never returns passwords or usernames.
user_ask	read	normal	Ask the person for a short piece of free text (a delivery address, a date, a preference) and return it. Never use it for passwords or codes; browser_login handles those.
user_choose	read	normal	Show a short list of options and return the person's choice. Use it instead of guessing when a decision belongs to them.
user_confirm	read	normal	Ask a yes/no question before something irreversible, such as paying or sending, and return the answer. Include the facts they need to decide (item, total, address).

learning (5)

Tool	Access	Risk	Description
learn_forget	write	write	Remove one learned lesson or preference by id (from learn_status). Use it when something MaskShift learned is wrong or no longer applies.
learn_status	read	normal	Show what MaskShift has learned from its own runs on this machine: lessons from earlier runs, preferences noticed from your messages, how each model and harness has done on which kinds of task, workflows it could turn into skills, and how well context was used.
router_explain	read	normal	Show how MaskShift would route a task: the models and the installed agent harnesses ranked by how they have actually done on similar tasks (success rate, cost), and how much evidence is behind each.
skill_mine	write	state	Look through recent runs for sequences of steps repeated across several runs that mostly went well, and draft a skill for each. Nothing is installed until skill_mine_accept.
skill_mine_accept	write	write	Install a skill drafted by skill_mine (by name or id), or dismiss it so it is not suggested again.

mcp (11)

Tool	Access	Risk	Description
mcp_add	write	install	Add any stdio or HTTP MCP server definition, including command, arguments, environment, URL, headers, and lazy loading settings.
mcp_call	write	external-tool	Invoke a connected MCP tool directly. Prefer activating its qualified tool so the model can call it natively on later turns.
mcp_connect	write	external-connect	Lazily start or connect an MCP server, negotiate modern or legacy protocol, and discover its tools.
mcp_disconnect	write	external-connect	Close a workspace-scoped MCP connection and unload its tool schemas.
mcp_list	read	normal	List curated, imported, project, and configured MCP servers with live connection status and tool counts.
mcp_prompts	read	normal	List prompt templates provided by an MCP server.
mcp_registry_install	write	install	Resolve a server from the official registry and add its remote or package transport to MaskShift configuration.
mcp_registry_search	read	normal	Search the live official MCP Registry instead of relying on a stale built-in server list.
mcp_resource_read	read	normal	Read a URI exposed by an MCP server.
mcp_resources	read	normal	Connect to a server and list its MCP resources.
mcp_search	read	normal	Search discovered MCP servers and tools already known to MaskShift. Servers remain lazy until connected.

memory (5)

Tool	Access	Risk	Description
memory_delete	write	write	Delete a memory by ID.
memory_list	read	normal	List memories ordered by effective importance: raw importance blended with a recency decay so stale, untouched memories sink without being deleted.
memory_optimize	write	write	Find duplicate-title memories to merge and stale, low-importance, never-accessed memories to prune. Defaults to a dry run that only reports candidates; set dryRun to false to apply the merge and prune.
memory_save	write	write	Save a durable project or global fact, architectural decision, convention, result, or reusable lesson. Automatically merges into an existing memory with the same title in the same scope instead of creating a duplicate, unless dedupe is set to false.
memory_search	read	normal	Search project and global long-term memory, ranked by a blend of text relevance, importance, and recency (older, untouched memories decay in rank without being deleted).

orchestration (16)

Tool	Access	Risk	Description
agent_cancel	write	agent	Cancel a running subagent or other active MaskShift run.
agent_dag_execute	write	agent	Execute the current dependency DAG in bounded parallel waves. Edit nodes default to isolated Git worktrees; failed dependencies block downstream work.
agent_delegate	write	agent	Run a focused subagent with its own session and capability context. Optionally give it a named persona (see agent_persona_list) to frame its system prompt, and optionally isolate editing in a Git worktree and branch.
agent_parallel	write	agent	Delegate multiple independent research, review, test, or implementation tasks concurrently and aggregate their final results. Each task may set its own persona.
agent_persona_list	read	normal	List named personas (bundled and project-defined) that can be passed as agent_delegate's persona argument to frame a subagent's system prompt for a specific role, such as reviewing code or resolving a build failure.
agent_route	read	normal	Recommend an available external coding-agent bridge or internal subagent based on the task profile.
agent_run_status	read	normal	Inspect active and recent agent runs, including parent/subagent relationships.
capability_activate	write	dynamic-load	Load selected capabilities into the current model context. Local tools add schemas, skills add instructions, and MCP servers connect lazily and expose their tools.
capability_search	read	normal	Search local tools, reusable skills, imported MCP servers, and discovered MCP tools. Use this whenever the current tool set is insufficient.
capability_state	read	normal	Show the exact tools, skills, and MCP servers currently loaded for this run.
model_calibrate	write	state	Run short probes (tool calling, a precise edit, planning, long-context recall) against a model and store the scores, so the harness sizes its help to what the model can actually do. Costs a few thousand tokens.
model_profile	read	normal	Show a model's scaffolding level (0 = leave it alone, 3 = carry it), why it was chosen, the knobs that level sets, and any stored calibration scores or observed track record.
model_route	read	normal	Rank configured models for a task using language/domain fit and prior MaskShift outcomes.
plan_dag_update	write	state	Create a dependency-aware execution plan whose ready nodes can run concurrently and whose dependent nodes receive predecessor results.
plan_get	read	normal	Return the current run plan and progress.
plan_update	write	state	Create or update the run plan with concise steps and statuses. Use it for multi-step work and keep it synchronized with actual progress.

plugins (7)

Tool	Access	Risk	Description
plugin_activate	write	host-exec	Load a plugin directly into the MaskShift process with full tool registration access.
plugin_deactivate	write	process	Run a plugin cleanup hook and remove its registered tools.
plugin_install	write	host-exec	Install a plugin from a local directory, Git repository, or npm package and activate it.
plugin_list	read	normal	List discovered runtime plugins, their status, tools, and skill directories.
plugin_reload	write	host-exec	Deactivate and re-import one plugin, or every discovered plugin, without restarting MaskShift.
plugin_scaffold	write	write	Generate and activate a complete single-tool MaskShift plugin scaffold.
plugin_scan	write	host-exec	Rescan user and workspace plugin directories and activate newly discovered plugins.

project (12)

Tool	Access	Risk	Description
change_impact	read	normal	Walk reverse imports and call relationships before an edit to identify affected symbols, modules, and likely tests.
code_graph_build	write	state	Build a persistent repository graph of files, symbols, imports, containment, and likely calls for dependency-aware navigation.
code_graph_query	read	normal	Find files and symbols in the persistent code graph, including incoming and outgoing relationship counts.
context_plan_explain	read	normal	Build a dry context plan and report budgets, selected source files, retrieval reasons, and excluded stale memories.
project_index_status	read	normal	Show local indexed file, chunk, character, and freshness statistics.
project_inspect	read	normal	Summarize repository shape, dominant languages, build manifests, instructions, Git state, and local index health.
project_instructions	read	normal	Load AGENTS.md, CLAUDE.md, MASKSHIFT.md, Copilot instructions, and other configured context files from repository root through the working directory.
project_read_manifest	read	normal	Read and optionally parse a package/build manifest such as package.json, pyproject.toml, Cargo.toml, go.mod, or CMakeLists.txt.
project_tree	read	normal	Render a bounded source tree for any workspace subdirectory.
provider_list	read	normal	Inspect configured model providers, connectivity status, and discovered models.
session_history	read	normal	Read messages and recent runs from the current MaskShift session.
usage_report	read	normal	Aggregate model token usage and estimated spend across recent runs, grouped by model, using the pricing table in config. Models without a configured price are reported as token counts only (priced:false), never guessed.

remote (2)

Tool	Access	Risk	Description
rsync_transfer	write	remote-exec	Synchronize local and remote files with rsync over SSH.
ssh_exec	write	remote-exec	Run a command on any SSH host with user, port, identity, jump host, environment, and timeout controls.

runtimes (2)

Tool	Access	Risk	Description
node_cell	write	host-exec	Execute arbitrary JavaScript as an ES module in a temporary script with workspace cwd.
python_cell	write	host-exec	Execute an arbitrary Python code cell in a temporary script with workspace cwd and return stdout/stderr.

search (6)

Tool	Access	Risk	Description
dependency_scan	read	normal	Inspect imports across selected source files and return a compact file-to-dependency graph without installing parsers.
repo_index	write	local-index	Build or refresh MaskShift’s local SQLite FTS code index for structure-aware context retrieval.
repo_search	read	normal	Search the local repository index using full-text ranking, blended with embedding-based semantic similarity when a local Ollama embedding model is available, and return bounded source chunks.
search_files	read	normal	Find files by fuzzy substring or glob across a workspace while respecting common repository ignores.
search_text	read	normal	Fast recursive source search using ripgrep when available, with regex, glob, case, hidden-file, and context controls.
symbol_outline	read	normal	Extract a lightweight outline of classes, functions, interfaces, structs, methods, and test blocks from source files.

shell (9)

Tool	Access	Risk	Description
command_lookup	read	normal	Find whether one or more commands are installed and return their executable paths.
shell_exec	write	host-exec	Run any Unix shell command on the host with the full user environment. Use for builds, tests, package managers, compilers, scripts, system inspection, and repository automation.
shell_exec_parallel	write	host-exec	Run multiple independent shell commands concurrently and return every result.
shell_process_list	read	normal	List processes launched by MaskShift, optionally limited to the current workspace or active processes.
shell_process_read	read	normal	Read current status and incremental output from a persistent process.
shell_process_stop	write	host-exec	Terminate a persistent process and its Unix process group.
shell_process_write	write	host-exec	Send text or control sequences to a persistent process stdin.
shell_start	write	host-exec	Start a persistent host process with live stdout/stderr streaming. Returns a process ID for later reads, input, or termination.
system_info	read	normal	Return operating system, CPU, memory, process, shell, runtime, and workspace information.

skills (7)

Tool	Access	Risk	Description
skill_create	write	write	Create a durable user skill from a successful workflow so future runs can discover it automatically.
skill_evaluate	write	state	Record comparable baseline and candidate outcomes for a reusable skill. This creates empirical evidence without changing the skill.
skill_improve	write	write	Append a validated lesson or refinement to an existing skill.
skill_load	read	normal	Load the full instructions and metadata for a selected skill into the current run.
skill_promote_validated	write	write	Apply a skill improvement only when recorded A/B trials meet minimum evidence and improve success without regressions.
skill_read_reference	read	normal	Read a file referenced by a skill while preventing path escape from the skill directory.
skill_search	read	normal	Search all bundled, project, Claude, Codex, Copilot, and user skill catalogs. Skill bodies are loaded only when selected.

storage (2)

Tool	Access	Risk	Description
storage_prune	write	state	Delete old checkpoints, stale search indexes, old run events and oversized logs according to the host-derived retention. Never deletes chats, memory or workspace files. Defaults to a dry run that only lists what would go.
storage_status	read	normal	Report how much disk MaskShift uses (database, checkpoints, browser profiles, logs), the budget derived from this machine's disk and memory, and advice when it is over budget.

system (7)

Tool	Access	Risk	Description
disk_usage	read	normal	Report filesystem-level free/used space (df), or a per-directory usage breakdown (du) when a path is given.
environment_list	read	secrets	List process environment variable names and optionally values. MaskShift autonomous mode permits direct secret-bearing environment access.
environment_set	write	secrets	Set or delete environment variables for this running MaskShift daemon and future child processes.
network_diagnose	read	normal	Run ping, DNS lookup, or traceroute against a host and return structured, parsed results instead of raw command text.
port_inspect	read	normal	Inspect listening sockets and processes using ss, netstat, or lsof.
ps_list	read	normal	List running processes on the host as structured records (pid, user, cpu%, mem%, command) — parsed, so nothing needs to eyeball raw ps output.
system_service	write	host-exec	Inspect, start, stop, restart, reload, enable, or disable a systemd service on the host.

web (3)

Tool	Access	Risk	Description
web_download	write	write	Download an HTTP(S) response directly to a host or workspace file, creating parent directories.
web_fetch	read	normal	Fetch any HTTP(S) URL with custom method, headers, and body. Returns bounded text, JSON, or readable text extracted from HTML.
web_search	read	normal	Search the public web through Brave, Tavily, or Exa when an API key is configured, falling back to DuckDuckGo HTML otherwise. For specialized search, activate an MCP search provider.