Daily
Generate or update workplace-facing daily reports from Tracework raw entries, using git only as limited fallback coverage. Supports work, personal, and private all-project scopes without mixing audiences. Use for "/tracework:daily", "更新日报", "写日报", "日报", "工作日志", "补日报", "daily note", or a dated daily work report. Do not use for agent handoff, meeting notes, generic prose reports, or git operations.From its SKILL.md
npx -y skills add KKenny0/Tracework --skill dailyAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
7.7 KB, ~1.7k tokens by cl100k_base, as published. Nobody here has run it
Tracework Daily
Write a concise daily management-closure view. Answer what changed today, why it matters, what is now closed or bounded, and the next gate. Do not produce a module inventory or agent handoff.
Required References
Read both before writing:
references/reporting-narrative-contract.md: scope partition, closure types, headline selection, evidence rules, and audience safety.references/daily-note-writing.md: source handling, merge rules, vault file shape, and conversation / first-run shapes.
Configuration
Resolve configuration in this order:
- Project
.tracework/config.yaml ~/.tracework/config.yaml
The primary output defaults to {vault}/Daily Note.md, overridable through
daily_note.path. Project paths come from daily_note.repos, then
{vault}/raw/projects.json, then the current repo.
If no vault can be resolved, return the report in the conversation. Do not create an unrequested vault. Do not block on cold-start. Optionally add one soft upgrade line at the end (see First-Run and Local Fallback).
First-Run and Local Fallback
Daily must produce value without prior setup. Cold-start is an upgrade for durable multi-day storage and strict audience partition, not a ticket to try.
Scope resolution
- If the user explicitly names
allor an exact reporting group such asworkorpersonal, that scope is explicit. - Else if
profile.default_reporting_groupis configured, that scope is configured. - Else the scope is implicit.
Partition rules
-
Explicit or configured
work/personal/ named group:- Include only projects whose
reporting_groupexactly matches. - Exclude
unassignedprojects. Never treat unassigned as safework. - If the result is empty because the current project is unassigned, say so
clearly, explain that explicit
work(or the configured default) will not invent a group, and tell the user how to fix it:/tracework:cold-start-interviewor setprofile.reporting_group. - Still return a short empty/excluded report rather than failing.
- Include only projects whose
-
Implicit scope (no explicit group, no configured default):
- If the current project has a
reporting_group, use that group. - If the current project is
unassignedor has no project config, enter local first-run:- Report only the current repository.
- Label scope
local(unassigned). Do not label itwork. - Use git and any available raw entries for that repo only.
- Mark evidence
limitedwhen git-only. - Add a one-line hint to set
reporting_groupbefore workplace-scoped reports.
- If a vault registry lists other assigned projects but the user is in an unassigned repo under implicit scope, prefer local first-run for the current repo and mention that assigned vault projects were not mixed in.
- If the current project has a
No vault
-
Always allowed.
-
Write nothing to disk.
-
Use the conversation output shape in
daily-note-writing.md. -
End with at most one upgrade line, for example:
可选:配置 knowledge vault 后可跨天累计并写入文件。
/tracework:cold-start-interview -
Do not imply the report is incomplete solely because vault setup is missing.
Empty signal
If local first-run has no raw entries and no meaningful git activity, return a short empty-state: what was checked, that nothing reportable was found, and that capture or more work signal will improve the next run. Do not error.
Workflow
-
Resolve date, scope class, and output target.
- Default date: today.
- Accept a date range and write one section per date.
- Resolve explicit / configured / implicit scope as above.
- Choose vault file output only when a vault exists and the resolved lane is
a normal scoped group (
work/personal/ named /allsections). Local first-run and no-vault runs stay in conversation even if a vault path exists, so unassigned content is not written into workplace Daily Note.
-
Partition projects before ranking.
- Read each project
profile.reporting_group, then itsprojects.jsonreporting_group. - Apply First-Run and Local Fallback together with the audience-safety rules in the shared narrative contract.
- For normal scoped reports, exclude unassigned projects and report the
missing classification. Show unassigned separately only in
all. - For local first-run, keep a single
locallane for the current repo.
- Read each project
-
Collect raw entries.
- Calculate the ISO week and read
{vault}/raw/weeks/{week}/{slug}.jsonfor the target date when a vault exists. - Prefer factual top-level fields and optional
reportingboundaries. - Artifact dossiers are optional navigation. Do not copy full artifacts.
- Without a vault, skip raw collection and continue with git coverage.
- Calculate the ISO week and read
-
Check git coverage.
- Use lightweight commit subject/stat inspection only for work not already covered by raw entries.
- Git-only material is
limited; do not infer motivation, decisions, verified impact, or management meaning that the commit does not support. - Filter formatting, generated-bundle, and chore-only noise unless it is a recorded risk or release gate.
-
Synthesize state transitions.
- Group entries and fallback commits by coherent work stream.
- Merge feature/fix/refactor entries that describe one state transition.
- For each reporting group or the single local lane, write one daily judgment and normally three headline advances. Two to four is acceptable when the evidence warrants it. Put remaining meaningful work under Other activity. Thin first-run reports may use fewer headlines.
- Preserve explicit risk, conflict, open question, rejected path, and evidence gaps.
-
Write incrementally or in conversation.
- Vault mode: match existing
### YYYY.MM.DDor### YYYY-MM-DDheadings; merge into an existing date section; never duplicate the date; preserve the exact project label- [Project Name]and stable field labels required by Monthly; do not write an empty date section. - No-vault or local first-run conversation mode: use the conversation shape
in
daily-note-writing.md. Do not invent a Daily Note path.
- Vault mode: match existing
Quality Gate
Before finishing, verify:
- Scope class (explicit / configured / implicit-local) was resolved before headline selection.
- A
workreport contains no personal or unassigned material, including evidence. - Local first-run is labeled
local/ unassigned, never presented as a safe workplaceworkreport. - Explicit scoped emptiness explains exclusion and repair, and does not leak
unassigned content into
work. - No-vault runs return conversation output without blocking on cold-start.
- Each reporting group or local lane has one clear daily judgment when there is signal.
- Headline items describe starting situation, movement, end state, meaning, and next gate when evidence supports them; git-only stays bounded.
- Remaining meaningful work is still covered when present.
ongoingandriskare not rewritten as completed outcomes.- The report is readable in about one minute per reporting group.
- Source/evidence boundaries are explicit.
References
references/reporting-narrative-contract.mdreferences/daily-note-writing.mdreferences/config-template.yaml
What ships with it: 5 files
16.0 KB alongside SKILL.md, 1 of them executable
agents/
- openai.yaml270 B
references/
- config-template.yaml2.2 KB
- daily-note-writing.md5.4 KB
- reporting-narrative-contract.md4.7 KB
scripts/
- git-stats.shruns3.5 KB
Gives 0 of the 12 instructions most docs writing skills give in ~1.7k tokens
Counted across 1,951 of the 3,904 authors here whose files we hold, read 2026-09-06
- Use third-person for skill descriptionsin 54 of 1951, across 35 files
- Start descriptions with Use whenin 43 of 1951, across 29 files
- Run baseline scenarios before writing any skillin 40 of 1951, across 26 files
- Use active voicein 40 of 1951, across 36 files
- Map file responsibilities before defining tasksin 36 of 1951, across 29 files
- Use checkbox syntax for tracking stepsin 35 of 1951, across 27 files
- Ask one question at a timein 35 of 1951
- Offer execution options after saving the planin 33 of 1951, across 24 files
- Include complete code in every stepin 33 of 1951, across 27 files
- Design units with clear boundaries and interfacesin 31 of 1951, across 23 files
- Announce the skill usage at the startin 30 of 1951
- Verify agent compliance after adding the skillin 29 of 1951, across 17 files
Said here and by no other author read
- resolve date scope and output target first
- partition projects by reporting group before ranking
- collect raw entries from vault if available
- use git coverage only for missing raw entries
- synthesize state transitions by coherent work stream
- write one daily judgment per reporting group
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.