Generate RLM-isolated follow-up messages enriched with chronicle events per user.
npx skills add https://github.com/0xHoneyJar/construct-observer --skill speaking
Please help me install this Agent Skill. Skill: speaking Source: https://github.com/0xHoneyJar/construct-observer/tree/main/skills/speaking Command: npx skills add https://github.com/0xHoneyJar/construct-observer --skill speaking
Generates RLM-isolated follow-up messages enriched with chronicle events around each user's last feedback date, enabling more contextually grounded outreach and better score movement explanations.
Run the /speak workflow to generate context-rich follow-ups for all active canvases using chronicle data.
High-intent search queries and answers about installing and using this skill.
You generate context-aware follow-up messages by running the workflow across per-user canvases, which queries chronicle events around each user's last feedback date to produce tailored, contextually grounded outreach.
Temporal context enrichment is the process of injecting chronicle event data around a user's last feedback date into generated outreach to explain score movements and provide contextually grounded messaging.
Yes, you need chronicle tooling, specifically the chronicle-query.sh script, and cognition data to run the follow-up generation workflow, as these components provide the temporal context and per-canvas readiness checks required.
Yes, you can batch generate follow-ups with per-user customization across multiple active canvases, using per-user cognition checks and chronicle data to tailor each message individually before delivery.
The process handles user readiness by performing cognition checks per user and per-canvas assessments to determine if the context is sufficiently grounded before delivering isolated follow-up messages.
Isolated follow-up messages for subagents require both chronicle data and cognition checks to function; without the necessary chronicle tooling or per-canvas cognition data, the workflow cannot ensure messages are contextually grounded.
argenAppend interaction details to a stakeholder's chronicle entry.
chekosGenerate sales outreach emails, negotiation scripts, and follow-up sequences from templates.
0xHoneyJarGenerates per-user cognition sidecars with provenance from canvas data.
HybridAIOneSummarize recent Discord messages and email threads into actionable catch-ups.
