Tools

Dynamo has built-in tools that the AI model uses to interact with your project. Run /tools to see the full list.

File Tools

read_file

Reads any file type. Dispatches by extension:

search_files

Powered by ripgrep (rg). Falls back to basic JS regex if rg is not installed.

Shell

run_command executes shell commands with:

• ●Two invocation modes — command: "string" (shell-wrapped, supports pipes/glob/redirect) or argv: ["array", "form"] (direct spawn, no escaping needed for paths with spaces). Exactly one must be provided.
• ●Persistent working directory — cd carries across tool calls. Reset on /clear.
• ●Configurable timeout — default 2 minutes, max 10 minutes. Set in dynamo.yaml under ai.shell.defaultTimeout.
• ●Background execution — long commands can run in the background. Model gets notified when they finish.
• ●Middle truncation — large output keeps the first and last 25K chars, truncates the middle.
• ●Shell detection — uses your $SHELL, not a hardcoded shell. On Windows it uses cmd unless SHELL names another; the model can ask for PowerShell.
• ●`DYNAMO=1` environment variable set so your scripts can detect they're running inside Dynamo.

Background Jobs

When the model runs a command with run_in_background: true:

1. 1.The command starts immediately and the model continues responding
2. 2.The jobs bar shows at the bottom: Jobs: 1 Ctrl+G to focus
3. 3.Press Ctrl+G to open the jobs picker with progress bars
4. 4.When the job finishes, the model auto-responds with the result
5. 5.Press Ctrl+K in the jobs picker to kill a running job

The model manages background jobs with these tools:

Web Tools

web_fetch

Fetches a URL, extracts article content with Readability, converts HTML to markdown.

• ●HTTPS upgrade — http:// URLs auto-upgraded to https://
• ●15-minute cache — same URL returns cached result
• ●Readability extraction — strips nav, ads, sidebars, returns article content only
• ●50K char cap — large pages truncated with notice
• ●30s timeout — prevents hanging on slow servers

web_search

Server-side search — the provider runs the search, not Dynamo. No separate API key needed.

Interactive

The model calls ask_user when it needs a decision — choosing between approaches, confirming a plan, selecting config options. Shows an interactive picker you navigate with arrow keys.

• ●Single or multi-question — up to 4 questions per call
• ●Tab navigation — switches between questions
• ●Option selection — arrow keys to navigate, Enter to select
• ●Custom input — "Other..." option opens inline text field for free-form answers

Task Management

Tasks track multi-step work with status and dependencies:

pending → in_progress → completed
                      → stopped

When a blocking task completes, dependent tasks auto-unblock. IDs are sequential — adding or deleting renumbers all tasks.

Use /tasks for an interactive task manager overlay, or let the model manage tasks via tool calls.

Git Tools

GitHub Tools

These need the GitHub CLI (gh) installed and authenticated. The seven list/view/diff tools are safe; the five that post something — gh_pr_create, gh_pr_comment, gh_pr_review, gh_issue_create, gh_issue_comment — are write.

The CI check rollup in gh_pr_view needs your GitHub token's "Checks: Read" permission. Without it, that one line says the rollup was denied and the rest of the PR still comes back. When a PR has no checks yet, it says none are reported instead of staying silent. gh_run_list and gh_run_view read Actions, a separate permission, and answer CI status either way.

Permissions

Tools are classified by risk level:

Any force-push to a protected branch — the default branch, a permissions.protected_branches pattern, or a GitHub-protected branch — is blocked in every spelling and argument order, by run_command and git_push alike. Set permissions.allow_force_push_protected: true in dynamo.yaml to make it ask every time instead. A plain push to a protected branch is dangerous, which Auto Mode approves.

"Allow every" is scoped per tool, per model, and persists across session resume. It never covers a dangerous call. Switching models doesn't wipe approvals — each model keeps its own grants.

--auto (-y) approves the safe, write and dangerous tiers as they come, so a task runs without step-by-step confirmation. Note that every run_command sits in the dangerous tier — ls as much as rm — so Auto Mode is one trust decision about the shell, not a per-command judgement. Unchanged by it: blocked patterns never run, and Dynamo still asks before it touches protected files like .env or keys, restores or deletes a checkpoint, stops a job, turns a skill back on, force-pushes to a protected branch you allowed, and during a compile-error cascade.

See Permissions & Security for the full security model.

Sandbox Mode

Controls filesystem access for all tools including shell commands.

Set via CLI (--sandbox project), config (ai.sandbox: project), or at runtime (/sandbox).

Checkpoints

Auto-checkpoint before file edits — snapshot the file's pre-write state so you can undo.

• ●Auto-checkpoint — created once per prompt, before the first write_file/edit_file and before destructive shell commands
• ●File-level snapshots — copies file content before modification. New files snapshot as "didn't exist"
• ●Rollback — restores files to pre-write state. Also rewinds conversation history.
• ●Disk persistence — survives restarts, stored under your config dir (per project)
• ●30-day cleanup — old checkpoints auto-purged on startup

Use /checkpoint for an interactive overlay to rollback, delete, or create manual checkpoints.

Configurable: ai.auto_checkpoint: false in dynamo.yaml to disable.

Syntax Highlighting

Tool results and model responses are syntax highlighted:

• ●Code blocks highlighted with Catppuccin Mocha theme (191 languages)
• ●File reads highlighted by file extension
• ●Diffs rendered with line numbers, background tints, and word-level inline diff
• ●Markdown rendered inline: headers, bold, inline code, links