clawock v0.3.1

[github-actions[bot]]

Install from PyPI:

python -m pip install clawock==0.3.1

A maintenance release for the Python package and clawock-dsh, with corrected
session accounting, clearer evidence and delivery readouts, and task-panel
improvements. Restart DeepSeek Harness after upgrading clawock-dsh so its
host methods and client bundle come from the same version.

Added

• clawock-dsh: /dispatch-list for OpenClaw, sharing the sidebar's
  provider balances, task queue, recent completions and patrol view (feat(dsh): /dispatch-list 改为任务芯片面板的文本版，复用 dsh 插件代码 #2168).
• clawock-dsh: task timelines, append history and patrol coverage, with
  readable findings, links to full task logs and Markdown chat layouts
  (feat: make patrol findings readable and clickable in dsh #2425, fix: give dsh task records an exit to the full log #2426, feat: expose task timelines and append history, fix recent completion ordering #2443, feat: expose patrol coverage and organize dispatch chat with Markdown #2518, feat: redesign the task chip's patrol coverage and the dispatch-list chat layout #2529).

Changed

• Dependency updates: Python now requires requests>=2.32.4 to include
  the credential-leak fix; the DSH protocol, client store and UI slots move
  to the 0.2.0-rc.2 line (build(deps): requests 下限抬到 2.32.4，跨过 CVE-2024-47081 (#2152) #2165, build(deps): bump the npm-safe group in /examples/dsh/packages/clawock-dsh with 4 updates #2578).
• The README and its diagrams explain the complete workflow before setup,
  with separate US/HK return curves, evidence and judgment feedback loops,
  and consistent icons and flow types (docs: the README reports the US and HK books apart, each with its own chart #2664, docs: lead the README with one complete loop figure on a single type scale #2735, docs: two return loops and the desk's maintenance loop in the overview; setup moves after the story #2736, docs: connectors in the README diagrams say what they carry — eight kinds of line, arrowhead and running icon #2740, docs: README diagram connectors carry what the edge moves, glyphs stop doubling up, the money arrow is one chevron #2749).

Fixed

• Portfolio accounting follows each market's trading session. Per-lot
  daily P&L handles same-session buys; malformed ledger rows are reported;
  unknown sessions fail the integrity check without crashing; quote refreshes
  preserve concurrent ledger edits and previous closes (fix: per-lot day P&L, malformed holding rows, intraday book commit #2355, fix: align HK quote sessions and dashboard return windows #2400, fix: require matching session for Today returns #2408,
  fix: report unknown holding sessions without crashing the money gate #2505, fix: ledger readers agree on rows, sessions and totals; a quote refresh keeps what was written during the fetch #2650). Leveraged bars are checked against their underlying where
  evidence is available, and suspect settlement marks carry refusal reasons
  (fix: settlement mark reasons, provenance check on rewritten rows, honest index day change #2357, fix: a leveraged fund's bar is graded with its underlying, not against a flat 50% line #2601).
• Lifecycle CLI flags match the selected phase and report mode, optional
  dependency errors name the required extra, and brief fallback honors the
  selected workspace (fix: commands that need an extra say which one #2554, fix: lifecycle --help offers only the flags a phase accepts; an empty commit is read from the index #2619, fix: readouts say what they count — watchdog entries vs passes, over-committed slots, a plan's own numbers, report flags per mode #2651, fix: honor the selected workspace in brief fallback #2734).
• Brief and intraday evidence checks use complete, dated sources. News
  windows follow publication time, stale-news checks cover all reference
  families, trigger text and currency totals are grounded in source data,
  and scorecard provenance detects rewritten rows (fix: anchor intraday news windows and check full live citations #2236, fix: derive brief book totals and verify currency arithmetic #2257, fix: settlement mark reasons, provenance check on rewritten rows, honest index day change #2357,
  fix: ground brief units and trigger text in their source data #2475, fix: the intraday stale-news gate reads the market-level reference families #2584, fix: four gate and path gaps (watchdog same-minute, live flashes, pending rows, cwd-relative workspace) #2598).
• Delivery and health checks preserve failures and recovery evidence.
  Watchdogs wait for an in-flight attempt, WeChat backstops deduplicate across
  checkouts, due slots with no record are counted, and rollups retain actual
  delivery timestamps and degradation reasons (fix: preserve publication failures and honor half-day report phases #2237, Delivery and security: backstop dedupe across checkouts, degraded verdicts, link scheme bypass, host paths #2314, fix: preserve scheduler and WeChat retry evidence #2483, fix: retain degradation reasons in the delivery summary #2487,
  feat: the delivery readout counts slots that were due and left no record #2602, fix: watchdogs hold a slot whose first attempt is still running; the WeChat retry verdict survives the heartbeat rewrite #2648, fix: dashboard reads the delivery time, the payload's today percentage and number-first weights; the sidecar fallback read is bounded #2649, fix: the scheduled-workflow rollup reads fresh runs, and the weekly review bounds its reasoning #2718).
• Dashboard tabs and financial readouts are reliable on first use.
  Inactive panels are inert; tab switching no longer waits on a failed load;
  negative amounts keep their sign; return windows, exposure bases, sample
  counts, source dates and holding sessions are labelled consistently
  (fix: make inactive dashboard panels inert at every viewport #2511, fix: disclose Reflect artifact dates and evidence build time #2513, fix: the first-paint Brier chip names its sample #2555, fix: dashboard quote sessions follow the market session, not the fetch clock #2581, fix: Reflect KPIs count market sessions once; brief holdings column names a change #2582, fix: dashboard readouts name their denominator, sample and basis #2583, fix: the hidden-concentration card declares its exposure basis and is checked against it #2600, fix: the first tab away from Overview no longer fails, hangs or waits on the network #2611, fix: the dashboard prints a negative amount one way #2697, fix: the gold card prints a loss with the minus outside the symbol #2719).
• clawock-dsh reports unreadable balances and queues as unavailable,
  expires trace caches, and re-renders changed FX provenance instead of
  retaining stale labels (dsh: Decision Mind says the ledger is unreadable instead of printing $0 (#2180) #2184, DSH plugin: balance and queue read honesty, chip border, npm readback by content #2319, fix: a stale quota reading stays yellow in the dsh panel row #2360, fix(dsh): unread balance is not healthy; debate arguments and FX provenance reach the panel #2586, fix: the dsh trace cache ages out, so the FX staleness line can appear #2659, fix(dsh): a changed FX provenance label re-renders Decision Mind #2696).
• Published Markdown rejects unsafe link targets and escapes model HTML;
  public workflow records redact host paths (Fix open patrol issues #2185–#2189, #2191 #2192, fix: block multiline unsafe Markdown link targets #2235, Delivery and security: backstop dedupe across checkouts, degraded verdicts, link scheme bypass, host paths #2314, fix: outcome ledger grouping, locking and redaction; credential scan on the data branch #2356).
• Release npm verification tolerates registry propagation and compares
  the served package with a fresh build in a separate readback job, allowing
  the GitHub Release after both registries accept publication (fix(release): npm 回读等住异步注册表，且不再挡 GitHub Release #2156, DSH plugin: balance and queue read honesty, chip border, npm readback by content #2319).

npm

clawock-dsh@0.3.1 published to npm.

---

This discussion was created from the release clawock v0.3.1.