| name | todo-txt |
| description | Use to track tasks, todos, checklists, plans, a work log, or any task list. Manages tasks using todo.txt — a plain-text format with priorities, contexts, tags, and due dates. |
| compatibility | opencode |
todo.txt Skill
Default file: todo.txt in the workspace root. done.txt for archived completed tasks. Plain text only. Create the file if it does not exist.
Format
One task per line. Keep each line short and atomic — one clear action. Split complex work into multiple tasks rather than writing a long description.
[priority] [creation-date] <description> [@context] [+tag] [key:value]
- Priority —
(A)–(Z) at position 0, followed by a space. Lowercase, mid-line, or missing space disqualifies it.
- Creation date —
YYYY-MM-DD immediately after priority+space, or at position 0 if no priority. Date elsewhere in the line is not a creation date.
- Contexts —
@word, Tags — +word. Both preceded by a space; appear anywhere after the priority/date prefix.
- key:value — structured metadata; both key and value are non-whitespace, non-colon strings. Common:
due:YYYY-MM-DD, pri:A, t:YYYY-MM-DD.
Completed tasks start with lowercase x at position 0, followed by completion date, then creation date (if present), then the original description:
x 2026-05-20 2026-05-01 Fix login bug @frontend +bug pri:B
X (uppercase), x mid-line, or x without a trailing space do not mark completion.
Sorting
The file MUST always be sorted. After every write, run:
sort todo.txt -o todo.txt
-o writes the result back in-place after reading is complete. Never use > todo.txt — it truncates the file before sort reads it. The format is designed so plain alphabetical sort produces the correct order: (A) < (B) < … < no-priority < x completed.
Operations
Get today's date with date +%Y-%m-%d (portable; date -I is a GNU/Linux shorthand for the same).
Add — compose the new line with today's creation date, append it to the file, sort.
Complete — read the file, find the line, strip only the leading (X) priority token if one is present (remove exactly (X) — two characters plus a space — and nothing else), prepend x YYYY-MM-DD (today). All remaining fields — creation date, description, contexts, tags, key:value pairs — stay in their original order. Optionally append pri:X to preserve the original priority. Write the file, sort.
Edit / Delete — read the file, modify or remove the line, write the file, sort.
Archive — run grep '^x ' todo.txt >> done.txt && grep -v '^x ' todo.txt > todo.txt.tmp && mv todo.txt.tmp todo.txt then sort todo.txt. The ^x pattern avoids false matches on words like xylophone.
Filter — grep for the token: @context, +tag (e.g. grep '+tech-debt' todo.txt), or ^(X) for a specific priority. To find tasks due on or before a date, grep for due: and compare values: grep 'due:' todo.txt; tasks are overdue if their due:YYYY-MM-DD value is earlier than today.
Example
(A) 2026-05-19 Call insurance @phone +urgent +home-repair
(B) 2026-05-01 Write unit tests @backend +tech-debt due:2026-05-31
2026-05-18 Read chapter 4 +learning +book
x 2026-05-20 2026-05-18 Submit expense report @admin +finance pri:B
x 2026-05-19 Fix login redirect @frontend +bug
