Summarization
Generate AI summaries of your daily notes, as well as summarize weeks worth of content.
Add two fields to swelog.json:
{ "llm": "ollama", "llmModel": "llama3.2"}See Providers and Models for supported providers and models.
Giving the model more context
Section titled “Giving the model more context”If you want to give the model more to work with, add a CONTEXT.md file in the
same folder as your work file — <obsidianVaultPath>/<swelogFolderName>/, next
to WORK.md.
Describe the long-term things your daily notes leave out — your role, the systems you own, what you are being measured on:
# Engineer Context
## Role and Team
Senior backend engineer on the Payments Platform team.
## Systems Owned
Checkout APIs, billing event pipeline, payout reconciliation jobs.
## Current Priorities
Improve billing reliability, reduce support escalations, and make deploymentssafer.Daily summaries
Section titled “Daily summaries”Daily summarization read the current work file and generate an AI summary of your notes with the original notes appended at the bottom.
swelog summarizeswelog summarize is an alias for swelog summarize day.
Weekly summaries
Section titled “Weekly summaries”swelog summarize week collects the summaries you generated on each day of the week and generates a weekly summary.
For example, if it the end of the day on Friday and you have 5 daily logs generated (Monday to Friday), the command will incorporate all of them into a weekly summary.
If it is the following Monday and you forgot to generate the weekly summary, you can use the --last-week flag to generate it for the previous week. See Backfilling a week for more details.