Free-interval session reports for DeepSeek Harness: AI-narrated daily/weekly digests over any date range + every-N-days scheduled reminders with system notifications
Pending human review — the facts below come from the source repository.
Free-interval session reports for DeepSeek Harness: AI-narrated daily/weekly digests over any date range + every-N-days scheduled reminders with system notifications
Users who want this capability in DSH; check the source and docs before installing.
dsh plugin --profile web add github:zhengjy01/dsh-period-report No supported platforms were declared.
Today at a glance / Highlights / Review / Suggestions) that reads like a human-written daily report. Falls back to a template highlight list when AI is disabled or unavailable. - Full statistics — active/new sessions, user & assistant messages, tool calls, turns, steps, token usage (input/output/cache read/write/reasoning), turn results, top tools. - Per-session details — title, id, working directory, active window, message/tool/turn counts, first request. - Periodic reminders — free period length: intervalDays: 1 = daily, 7 = weekly, 2 = every other day… anchored at anchorDate, fired at hour:minute, once per period. range: previous reports the completed previous period (e.g. 09:00 digest of yesterday); range: current reports the period in progress. - System notifications — macOS Notification Center banner via osascript, Linux via notify-send. Full reports are also saved to ~/.dsh/dsh-period-report/reports/ so nothing is lost when the GUI is closed. - Bilingual — reports in Chinese (zh) or English (en). - Persistent config — ~/.dsh/dsh-period-report.json (mode 0600). ## Install bash dsh plugin --profile web add github:zhengjy01/dsh-period-report Local development: bash dsh plugin --profile web add link:/path/to/dsh-period-report ## Tools ### report_generate Generate a report for an arbitrary date range. | Parameter | Type | Description | | --- | --- | --- | | startDate | string (required) | Start date, YYYY-MM-DD (inclusive) | | endDate | string (required) | End date, YYYY-MM-DD (inclusive) | | includeSubagents | boolean | Include subagent sessions (default false) | | withUsage | boolean | Include token usage stats (default true; disable for speed) | | language | string | zh or en (default zh) | | narrate | boolean | AI narrative digest (default true; disable for a fast template digest) | Returns the full Markdown report. ### report_config View or update the reminder configuration (persisted in ~/.dsh/dsh-period-report.json). | Parameter | Type | Description | | --- | --- | --- | | enabled | boolean | Enable/disable the scheduled reminder | | anchorDate | string | Anchor date YYYY-MM-DD; periods count from this day | | intervalDays | number | Remind every N days (1–90; 7 = weekly) | | hour / minute | number | Reminder time of day | | range | string | previous (report the last period, default) or current | | language | string | Report language zh / en | | narrate | boolean | AI narrative toggle | Calling it with no arguments returns the current configuration. ## Example > “Set a reminder every 2 days at 09:00 starting from 2026-06-01, in Chinese.” > > → report_config with { "enabled": true, "anchorDate": "2026-06-01", "intervalDays": 2, "hour": 9, "minute": 0, "range": "previous" } > > Reminders fire on 06-01, 06-03, 06-05 … at 09:00, each reporting the previous 2-day period; a system notification shows the summary and the report file path. ## License MITNot by stars — picked by hand, one at a time.
A visual settings dashboard for DeepSeek Harness: every registered settings namespace — including third-party plugin ones the official UI never covers — rendered as editable, schema-driven forms. Early, but the right direction: stop hand-editing YAML.
A DIY trajectory visualizer for DeepSeek Harness: watch the agent’s plan and execution unfold as a graph instead of scrolling raw logs. Early, but it fills a real gap — understanding at a glance what the agent is doing.
Dispatch DSH work from Claude Code / Codex: spin up in-host DSH agent sessions with per-tier presets, watch native subagent progress, and borrow a multimodal bridge that lends the text-only harness vision and image generation. An orchestration plugin that turns DSH into a backend for other coding agents.