delonet-daily-report

name

delonet-daily-report

description

Operate the DeLoNET Daily Report pipeline; configure tracked news topics and sources, batch three ephemeral investigator roles, reconcile durable Hermes cron journalists named ddr:journal:{id} and the ddr:daily aggregator, validate SectionArtifact/RunManifest/DailyReport JSON contracts, compose freshness-aware briefs and a daily company rollup, archive reports, and diagnose scheduler health. Use for reportctl, daily intelligence reports, requests like "add X to my daily report", topic add/update/pause/resume/remove, stale or duplicate sections, source policy, journalist lifecycle, cron reconciliation, and report safety. Do not use for Hermes fleet provisioning, Bloodbank changes, general news writing, or arbitrary cron administration.

DeLoNET Daily Report

Route daily-report work to the smallest necessary reference set, then use scripts/reportctl for deterministic state changes and reconciliation.

Operating principles

• Keep configuration as the source of truth; treat Hermes cron as reconciled runtime state.
• Pin inference.provider and inference.model to the authoritative active Hermes profile; recreate jobs whose native snapshots drift.
• Use three ephemeral investigator roles per topic run when subagents are available; never make investigators durable.
• Make journalists durable only as ddr:journal:<id> jobs and keep exactly one ddr:daily aggregator.
• Preserve evidence lineage, timestamps, and freshness state through every contract boundary.
• Plan first. Apply cron changes only after reviewing the stable reconciliation plan.

Route by intent

Need to change the system?
├─ Understand components, paths, or contracts → references/architecture.md
├─ Add or assess feeds, sites, APIs, or trust rules → references/sources.md
├─ Run multi-agent research for one topic → references/investigator-workflow.md
├─ Add/update/pause/resume/remove topics or reconcile cron → references/journalist-lifecycle.md
├─ Build, validate, or archive a daily report → references/report-composition.md
└─ Diagnose duplicates, stale sections, leaked secrets, or unsafe output → references/safety-gotchas.md

Common combinations

Scenario	Read in order
Bootstrap or redesign the pipeline	references/architecture.md → references/journalist-lifecycle.md
Add a topic with credible sources	references/sources.md → references/journalist-lifecycle.md
Execute a journalist run	references/investigator-workflow.md → references/report-composition.md
Debug a bad daily brief	references/safety-gotchas.md → references/report-composition.md
Recover scheduler drift	references/journalist-lifecycle.md → references/safety-gotchas.md

Controller workflow

1. Copy assets/example-config.json outside the skill and customize it.
2. Run scripts/reportctl --config PATH validate.
3. Mutate topics with topic add|update|pause|resume|remove; writes are atomic.
4. Run scripts/reportctl --config PATH plan --jobs SNAPSHOT.json for offline planning, or omit --jobs to read canonical $HERMES_HOME/cron/jobs.json.
5. Review commands, then run scripts/reportctl --config PATH reconcile --apply. Never use --apply casually.
6. Archive validated output with archive --report REPORT.json --markdown REPORT.md; writes are atomic.
7. Run status, health, and paths --date YYYY-MM-DD to inspect coverage and archive locations.

Cross-cutting rules

• Never embed credentials. Store only public source URLs and environment-variable names; redact secret-like values from status and plans.
• Reject silent data loss. Mark missing, invalid, or old section artifacts stale; do not omit them without a manifest entry.
• Use stable identifiers. Topic IDs are lowercase kebab-case and permanently determine job names and section paths.
• Keep aggregation separate. Journalists emit SectionArtifact; only ddr:daily emits RunManifest and DailyReport.
• Batch investigators. Spawn the primary-source researcher, change tracker, and skeptic concurrently in one request when delegation exists; collect results, then terminate them.

Out of scope

This skill covers one report pipeline and its owned cron jobs. It does not cover:

• Hermes installation, profiles, or fleet provisioning: use agent-fleet-operations.
• Bloodbank event schemas or publishers: use bloodbank-integration; make no Bloodbank edits.
• General-purpose cron administration: use Hermes cron documentation directly.
• n8n workflow design: use delonet-n8n-architecture.
• Live configuration installation: operators choose paths and explicitly apply changes; this package never writes ~/.config or ~/.hermes during setup or tests.