log

Solid

the shape of a daily agent log: threads carrying their own notes and prompts (saves log to .construct/)

AI & Automation 3 stars 0 forks Updated today MIT

Install

View on GitHub

Quality Score: 82/100

Stars 20%
20
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
100
Issue Health 10%
80
License 10%
100
Description 5%
100

Skill Content

**today's work, shaped for tomorrow's session:** the next agent reads it instead of asking you - threads group work by topic, carrying their own notes and prompts - `inject-log` carries the four most recent threads forward across days - the stop hook demands it, so a turn cannot close on an unwritten day # .construct/retardify/log/YYYY-MM-DD.md one file per day, holding both the work and the prompts that drove it: - gitignored in this repo; host projects decide for themselves whether to track it - `logs` are written in MAXIMALLY clear, concise, casual language, skipping trivial details - `threads` group work by task or topic, limited to 50 lines of prose, prompts excluded - a thread also holds to 5000 bytes whole, prompts included, since sessionstart injects it entire - that cap is why nothing truncates a log: the writer spends the budget, the reader never guesses - sessionstart carries the 4 most recent threads forward, taken across days rather than per file - `sections` lead with the main idea, followed by supporting ideas - `lines` hold a single clause, fact or action, capped at 100 characters - `notes` are appended after taskcomplete or every 30 minutes, limited to 5 bullets - `prompts` are appended to the thread they drove, rewritten short, always timestamped - a prompt that starts a new thread belongs to the thread it created, not the one before - no trigger wraps it: a thread, a note or a synthesis is asked for in plain words - minimize comma chains, em dashes, **bol...

Details

Author
MaisonDeVolonte
Repository
MaisonDeVolonte/construct
Created
1 months ago
Last Updated
today
Language
Shell
License
MIT

Similar Skills

Semantically similar based on skill content — not just same category

Data & Documents Listed

log

Writes one day's work log as a file — what moved in the tracker, what was on the calendar, what was decided, what other people said about your work, and links to the evidence. Built to run unattended on a schedule, so it never asks and never invents; a source it cannot reach is reported as missing rather than filled in. It also looks back a few days and fills the gaps an earlier run left. Turning these files into accomplishment statements is /pm:log-review. Triggers - explicit invocation only - "/pm:log", "write the log for today", "fill in the days that were missed", "일지 써줘", "업무 일지 작성", "빠진 날 채워줘".

5 Updated today
byjunyoung
AI & Automation Listed

log-session

Append a structured entry to the project's session log (`docs/LOGS.md`): what was done this session, files touched, decisions taken, and the next step — so the next session (or another person) can pick up the thread without re-reading git history. Run it before `/clear`, before closing Claude Code, or at any natural stopping point. On Claude Code and want hand-tuned per-skill model/effort tiers? Install the `#claude` branch instead (`npx skills add gtrabanco/agentic-workflow#claude`) — see the README. This branch is model-agnostic: the skill inherits whatever model and effort your agent session is already using. Triggers: "log this session", "log-session", "write a session log", "journal what we did", "record this session before I clear", "save the session summary".

21 Updated today
gtrabanco
AI & Automation Listed

proglog

Append a dated entry to the user's private pairing-journal file — concepts relearned (with an analogy to a prior stack when one genuinely fits) and what got built this session. Invoke via /proglog, or proactively offer it at the end of a substantive hands-on pairing/relearning session. Also proactively check at the START of a session that looks like a continuation of pairing work: if the latest prog-log entry looks like it predates work that's evidently happened since, remind the user before diving back in. Do NOT use for quick one-off questions, mechanical edits, or sessions with no relearning/building content worth journaling — and don't nag if the user declines the reminder once.

0 Updated 4 days ago
sanlee-ys