daily-recap
Daily Recap
End-of-day consolidation. Triggered by heartbeat at 21:00-22:00 or manually ("recap the day").
Algorithm
1. Read today's log
Open memory/YYYY-MM-DD.md. If empty or no entries — write "No entries today" and HEARTBEAT_OK.
2. Extract key items
From all entries of the day, compile:
## Daily Recap
**Decisions:** (decisions made today)
- [decision]: [why] -> [what changes]
**New facts:** (new confirmed facts)
- [fact] (source: [where from])
**Belief changes:** (what shifted in understanding)
- [was] -> [became] (evidence: [what showed it])
**Open threads:** (started but not finished)
- [topic]: [what remains to be done]
**Constraint check:** [current constraint] — today [exploit/subordinate/ignore]?
3. Check for promotion to MEMORY.md
For each new fact and belief change, ask:
- Does this change something in MEMORY.md sections? (Company, Wedge, PMF Stage, Metrics, Constraint)
- Is it sufficiently confirmed? (not a single data point)
If yes — update MEMORY.md:
- Replace outdated info (don't duplicate)
- Update review timestamp if Constraint or PMF Stage changed
- Keep MEMORY.md < 3KB
4. Write recap to daily log
Append the ## Daily Recap block to the end of memory/YYYY-MM-DD.md.
5. Update heartbeat state
In memory/heartbeat-state.json set "lastRecapDate": "YYYY-MM-DD".
Telegram Output Format
Day recap:
Decisions: X
- brief description of each
New facts: Y
- brief description of each
Open: Z threads
[list if any]
Constraint: [exploit/subordinate/ignore today]
If the day was empty: "Quiet day. No entries."
Follow output preferences from USER.md (language, format, platform constraints).
Rules
- Max 15 lines for Telegram
- Don't fabricate — only from today's entries
- Belief changes > facts > tasks by importance
- If MEMORY.md was updated — mention: "MEMORY.md updated: [what exactly]"
More from codealive-ai/ai-cofounder
exa-code-search
Code search using Exa. Finds real code snippets, API docs, and examples from GitHub, StackOverflow, and technical docs. Use when searching for code examples, API syntax, library documentation, or debugging help.
2exa-people-research
People research using Exa search. Finds LinkedIn profiles, professional backgrounds, experts, team members, and public bios. Use when searching for people, finding experts, or looking up professional profiles.
2strategic-review
Deep multi-phase strategic review of a specific decision — premise challenge, alternatives, risk map, adversarial review. Use for big choices (wedge, pricing, channel, hire, pivot). For quick framework lookup, use decision-playbook. For diagnosing why something is broken, use business-investigation.
2enterprise-sales
Help users navigate enterprise sales. Use when someone is closing large deals, managing complex buying committees, handling procurement, or converting PLG users to enterprise contracts.
2web-browser
Browse the web using agent-browser CLI with Lightpanda engine. Read pages, extract content, interact with elements.
2launch-marketing
Plan and execute general product launches — PR outreach, GTM campaigns, buzz generation. For Product Hunt specifically, use product-hunt-launch (specs/tactics) and ph-launch-timeline (30-day prep plan) instead.
2