Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add furkanzt/claude-mstodo/plugin install mstodoWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/furkanzt/claude-mstodo/mstodo)<a href="https://agentmods.dev/commands/furkanzt/claude-mstodo/mstodo"><img src="https://agentmods.dev/badge/commands/furkanzt/claude-mstodo/mstodo/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/furkanzt/claude-mstodo/mstodo"><img src="https://agentmods.dev/badge/commands/furkanzt/claude-mstodo/mstodo.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00019 | $0.02962 |
| Opus 5 | $0.00010 | $0.01481 |
| Sonnet 5 | $0.00004 | $0.00592 |
| Haiku 4.5 | $0.00002 | $0.00296 |
Grade A, and why
mstodo scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.
The user invoked /mstodo to work on their Microsoft To Do tasks via the mstodo MCP tools (todo_login, todo_get_lists, todo_get_tasks, todo_create_task, todo_complete_tasks, todo_update_task, todo_add_note, todo_checklist, todo_set_groups).
Requested list/group: "$ARGUMENTS"
Follow this flow:
-
Install self-check — first
/mstodoof the conversation only. If you have already run this check earlier in this conversation, skip straight to step 1; do not re-check, and do not re-report a problem you already reported. On the first invocation, run the single Bash call in Install self-check below. If everything is fine, say nothing about it and carry on silently — this check is invisible when healthy. If it reports a problem, tell the user in one or two lines, then carry on with the request anyway using whatever still works. Never repair it on your own initiative. Do not go looking for where the repo moved to, do not rewrite the symlink, do not editcli-path.json— not even if you can see the correct path. Say what is broken and what would fix it, then wait to be asked. A fresh conversation re-runs the check, and the user can also just ask you to check again. -
reset(also accept--clear-default) — check for this before parsing the text as a list name. Delete this project's entry from~/.mstodo-mcp/defaults.json(see Per-project default list below), confirm in one line — "Cleared the default list for this project." — and stop. Leave every other project's entry untouched. -
No argument given (the text above is empty): read
~/.mstodo-mcp/defaults.jsonand look up the current working directory.- A default is stored → use it immediately, without asking. But announce it in one line first, before any tool call:
Using your default list for this project: **Backlog** (Work). Then continue at step 4 with that list (and group). Never silently load a list. - No default, or the file is missing/unreadable → call
todo_get_lists, show the lists grouped by folder (and any ungrouped ones), and ask which list to work on. Then stop and wait. Once the user names a list, fetch it — and then ask once: "Make X the default list for this project?" Write it todefaults.jsononly if they say yes.
- A default is stored → use it immediately, without asking. But announce it in one line first, before any tool call:
-
Argument given: interpret it as a list name that may include a group. Accept natural phrasings — "Backlog from Work", "Work / Backlog", "Work: Backlog", or just "Backlog". Extract the group when present and call
todo_get_taskswithlist(andgroupif identified). Defaults to open tasks only, which is what you want. An explicit argument wins for that invocation and never changes the stored default — not even when it names a different list, not even when it's used all session. The user often works two or more lists at once for different purposes. The stored default is only what a bare/mstodoreaches for, and it changes only when the user explicitly asks. -
Handle the tool's response:
needs_clarification(ambiguous or not found): show the returned candidates/suggestions and ask the user which they mean, then retry with the group set (orlist_id). Do not guess when two lists share a name.not_signed_in: calltodo_login, relay the sign-in URL and code to the user, and wait for them to confirm they've signed in before retrying. Don't loop silently.
-
If any returned task has
notesTruncated: true— its note was cut at 1000 characters by the MCP server. Fetch the full text immediately, before you present that task, using the read-only companion CLI (see Reading full notes below). Do not summarize or plan around a truncated note, and do not ask permission first — the user would rather always see everything than risk missing something buried in a long note. If several tasks are truncated, fetch them all in one Bash call. -
Once tasks are fetched, present them clearly — group by status, surface due dates, importance, notes (full text, per step 5) and checklist subtasks. Then plan together: help prioritize, break big items into checklist subtasks (
todo_checklist), set or adjust due dates/importance (todo_update_task), or create new tasks (todo_create_task) as the conversation calls for it.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 119 lines · 0 tokens per session scan A 28070e9d9fb2
mstodo is a command published in the GitHub repository furkanzt/claude-mstodo (0 stars, last pushed 1mo ago), licensed MIT. It adds 19 tokens to every session and 2,962 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
standup
Quick daily standup from today's Claude Code sessions — grouped by project, with cost and errors.
fire-todos
Capture, list, and manage todos during work sessions.
gsd:resume-work
Resume work from previous session with full context restoration.
todo.template
This prompt was authored for Claude-style slash workflows. In Codex runtime, adapt tool calls as follows.
ops
Weekly operations loop — triage the inquiry inbox, advance the CRM pipeline, sweep open PRs across repos, and surface the content plan. Loads the ops skill.
standup
A command that creates a daily stand-up report by summarizing recent work, planned tasks, calendar events, and blockers.