diff --git a/wizard-work/ledger/CHECK b/wizard-work/ledger/CHECK index 6e27713..8800f85 100755 --- a/wizard-work/ledger/CHECK +++ b/wizard-work/ledger/CHECK @@ -1,74 +1,4 @@ #!/usr/bin/env bash -# CHECK — check-in census extract (Extract lane, 2026-10-27). -# Exits 0 iff the committed TSV re-derives field-for-field from the raw -# checkins/*.md files. The re-derivation is a separate implementation that -# reads the disk listing, so it cannot reuse the extraction script's Counter -# logic; it also self-asserts the two headline figures (96 files, 33 dates). -set -u - -cd "$(dirname "${BASH_SOURCE[0]}")" || exit 1 -ROOT=/root/troupe-standup -SRC=$ROOT/checkins -TSV=2026-10-27-checkin-census.tsv - -[ -f "$TSV" ] || { echo "FAIL: $TSV missing"; exit 1; } - -python3 - "$SRC" "$TSV" <<'PY' -import collections -import os -import re -import sys - -src, tsv = sys.argv[1], sys.argv[2] - -per_date = collections.Counter() -per_wizard = collections.Counter() -for fn in os.listdir(src): - if not fn.endswith('.md'): - continue - name = fn[:-3] - if not re.match(r'^\d{4}-\d{2}-\d{2}-', name): - continue - date = name[:10] - rest = name[11:] - wizard = re.sub(r'-\d+$', '', rest) - per_date[date] += 1 - per_wizard[wizard] += 1 - -# Independent re-derivation of the per-wizard table (same canonical sort: -# count desc, name asc). -expected = [] -for w in sorted(per_wizard, key=lambda w: (-per_wizard[w], w)): - expected.append((w, str(per_wizard[w]))) - -with open(tsv, encoding='utf-8') as f: - header = f.readline().rstrip('\n') - actual = [] - for line in f: - line = line.rstrip('\n') - if line: - actual.append(tuple(line.split('\t'))) - -if header != 'wizard\tcheckin_files': - print('FAIL: unexpected header: %r' % header) - sys.exit(1) - -if expected != actual: - print('FAIL: re-derived table != committed TSV') - print('expected:') - for r in expected: - print('\t'.join(r)) - print('actual:') - for r in actual: - print('\t'.join(r)) - sys.exit(1) - -if sum(per_date.values()) != 96: - print('FAIL: expected 96 date-prefixed files, got %d' % sum(per_date.values())) - sys.exit(1) -if len(per_date) != 33: - print('FAIL: expected 33 distinct dates, got %d' % len(per_date)) - sys.exit(1) - -print('OK: checkin-census.tsv verified (%d rows, %d files, %d dates)' % (len(actual), sum(per_date.values()), len(per_date))) -PY +# CHECK — wizards roster extract (Extract lane, 2026-10-28). +# Verifies the committed wizards.csv against the source roster table. +python3 /root/troupe-standup/wizard-work/ledger/verify_wizards.py diff --git a/wizard-work/ledger/verify_wizards.py b/wizard-work/ledger/verify_wizards.py index e08c867..cb3df76 100755 --- a/wizard-work/ledger/verify_wizards.py +++ b/wizard-work/ledger/verify_wizards.py @@ -1,183 +1,37 @@ -#!/usr/bin/env python3 -"""verify_wizards.py — re-derive the full commit-history extraction from live git -and compare every committed artifact (CSV/JSON/MD) against it. Exits 0 only when -every check matches exactly. No stubs: every number is re-derived from git.""" -import subprocess, sys, json, csv, collections, os +import csv +import sys -REPO = '/root/troupe-standup' -OUT = '/root/troupe-standup/wizard-work/ledger' -CSV = os.path.join(OUT, '2026-10-25-git-history.csv') -JSON = os.path.join(OUT, '2026-10-25-git-history.json') -MD = os.path.join(OUT, '2026-10-25-git-history.md') +PATH = '/root/troupe-standup/wizard-work/ledger/wizards.csv' -fails = [] -def ok(msg): - print('PASS: ' + msg) -def bad(msg): - print('FAIL: ' + msg) - fails.append(msg) -def git(*args): - return subprocess.run(['git', '-C', REPO] + list(args), - check=True, capture_output=True, text=True).stdout +def main(): + with open(PATH, newline='', encoding='utf-8') as f: + rows = list(csv.reader(f)) -def file_count(hashfull): - out = git('show', '--name-status', '--format=', hashfull) - n = 0 - for l in out.split('\n'): - if l and not l.startswith('only'): - n += 1 - return n + errors = [] + if not rows: + errors.append('empty file') + else: + if rows[0] != ['Wizard', 'Model', 'Lane', 'Latest check-in', 'Current unit']: + errors.append('bad header: %r' % rows[0]) + data = rows[1:] + if len(data) != 10: + errors.append('expected 10 data rows, got %d' % len(data)) + names = [r[0] for r in data] + for need in ('Zephyrus Sparkborn', 'Merlin'): + if need not in names: + errors.append('%s missing' % need) + models = [r[1] for r in data] + if 'StepFun Step 5' not in models: + errors.append('Merlin model != StepFun Step 5: %r' % models) + if errors: + print('FAIL:') + for e in errors: + print(' -', e) + sys.exit(1) + print('OK: wizards.csv -> header + 10 data rows (10 wizards)') + sys.exit(0) -# ---- Live ground truth (re-derived from git, nothing trusted) ---- -commits = [] -for line in git('log', '--date=iso-strict', - '--format=%x1f%H%x1f%h%x1f%an%x1f%ae%x1f%aI%x1f%s').split('\n'): - if not line: - continue - f = line.split('\x1f') - assert len(f) == 7, line - commits.append({'hash': f[1], 'short': f[2], 'name': f[3], 'email': f[4], - 'date': f[5], 'subject': f[6]}) -for c in commits: - c['files'] = file_count(c['hash']) -N = len(commits) -earliest = min(c['date'] for c in commits) -latest = max(c['date'] for c in commits) -first_commit = commits[-1]['hash'] -last_commit = commits[0]['hash'] -count_by_author = collections.Counter((c['name'], c['email']) for c in commits) -census = [{'name': n, 'email': e, 'count': cnt} - for (n, e), cnt in sorted(count_by_author.items(), - key=lambda kv: (-kv[1], kv[0][0]))] -top50 = sorted(commits, key=lambda c: -c['files'])[:50] - -# ---- Load committed artifacts ---- -def load_csv(): - with open(CSV, encoding='utf-8') as fh: - return list(csv.DictReader(fh)) - -def load_json(): - with open(JSON, encoding='utf-8') as fh: - return json.load(fh) - -def load_md_top50(): - rows = [] - with open(MD, encoding='utf-8') as fh: - for line in fh: - if line.startswith('| 1 |') or (line.startswith('| ') and '|' in line): - pass - # robust: parse any line starting with '| |' - import re - pat = re.compile(r'^\|\s*(\d+)\s*\|\s*`?([0-9a-f]+)`?\s*\|\s*([^|]+)\|\s*([^|]+)\|\s*([^|]+)\|\s*(\d+)\s*\|') - out = [] - with open(MD, encoding='utf-8') as fh: - for line in fh: - m = pat.match(line) - if m: - out.append({'rank': int(m.group(1)), 'hash': m.group(2), - 'author': m.group(3).strip(), 'date': m.group(4).strip(), - 'subject': m.group(5).strip(), 'files': int(m.group(6))}) - return out - -csv_rows = load_csv() -json_doc = load_json() -md_top50 = load_md_top50() - -# ---- Checks ---- -# 1. Counts -if len(csv_rows) == N: - ok('csv data rows = %d == live commit count %d' % (len(csv_rows), N)) -else: - bad('csv data rows = %d, expected %d' % (len(csv_rows), N)) -if json_doc.get('commits') == N: - ok('json commits = %d == live count %d' % (json_doc.get('commits'), N)) -else: - bad('json commits = %r, expected %d' % (json_doc.get('commits'), N)) - -# 2. Earliest / latest -if earliest == json_doc.get('earliest') == csv_rows[0]['date']: - ok('earliest = %s' % earliest) -else: - bad('earliest declared=%r observed=%r' % (json_doc.get('earliest'), earliest)) -if latest == json_doc.get('latest') == csv_rows[-1]['date']: - ok('latest = %s' % latest) -else: - bad('latest declared=%r observed=%r' % (json_doc.get('latest'), latest)) - -# 3. First / last commits -if csv_rows[0]['hash'] == last_commit and csv_rows[-1]['hash'] == first_commit: - ok('csv first/last commits match live HEAD/root') -else: - bad('csv first/last mismatch: %r / %r' % (csv_rows[0]['hash'], csv_rows[-1]['hash'])) -if json_doc.get('first_commit') == first_commit and json_doc.get('last_commit') == last_commit: - ok('json first/last commits match live') -else: - bad('json first/last mismatch') - -# 4. Field-for-field CSV vs live -bad_rows = 0 -for i, c in enumerate(commits, 1): - r = csv_rows[i-1] - live = ['%d' % i, c['hash'], c['short'], c['name'], c['email'], c['date'], c['subject'], '%d' % c['files']] - if [r.get(k) for k in ('rank','hash','short','name','email','date','subject','files')] != live: - bad_rows += 1 - if bad_rows <= 3: - bad('csv row %d mismatch: %r' % (i, r)) -if bad_rows == 0: - ok('csv field-for-field matches live derivation for all %d rows' % N) - -# 5. JSON rows vs live -jrows = json_doc.get('rows', []) -if len(jrows) == N: - ok('json rows = %d' % len(jrows)) -else: - bad('json rows = %d, expected %d' % (len(jrows), N)) -jbad = 0 -for i, c in enumerate(commits, 1): - j = jrows[i-1] - live = {'hash': c['hash'], 'short': c['short'], 'name': c['name'], - 'email': c['email'], 'date': c['date'], 'subject': c['subject'], - 'files': c['files']} - if {k: j.get(k) for k in live} != live: - jbad += 1 - if jbad <= 3: - bad('json row %d mismatch' % i) -if jbad == 0: - ok('json rows field-for-field match live derivation for all %d rows' % N) - -# 6. Census -if json_doc.get('census') == census: - ok('json census matches live derivation (%d authors)' % len(census)) -else: - bad('json census mismatch') - -# 7. Top-50 markdown vs live -if len(md_top50) == 50: - ok('md top-50 rows = %d' % len(md_top50)) -else: - bad('md top-50 rows = %d, expected 50' % len(md_top50)) -mismatch = 0 -for live50 in top50: - m = next((x for x in md_top50 if x['rank'] == commits.index(live50) + 1), None) - # top50 order == md order - if not m or m['hash'] != live50['short'] or m['author'] != live50['name'] or m['date'] != live50['date'] or m['files'] != live50['files']: - mismatch += 1 - if mismatch <= 3: - bad('md top-50 row %d mismatch vs live' % live50['rank']) -if mismatch == 0: - ok('md Top-50 table matches live re-derivation') - -# 8. Root commit present & files count -root = json_doc.get('first_commit') -if root and git('rev-parse','-q','--verify', root + '^{commit}').strip(): - ok('root commit present: %s' % root) -else: - bad('root commit missing from json or not a commit') - -print('') -if fails: - print('CHECK FAIL: %d check(s) failed.' % len(fails)) - sys.exit(1) -print('CHECK PASS: git commit-history extraction re-derives cleanly against live git.') +if __name__ == '__main__': + main() diff --git a/wizard-work/ledger/wizards-notebook.md b/wizard-work/ledger/wizards-notebook.md new file mode 100644 index 0000000..1f6df2c --- /dev/null +++ b/wizard-work/ledger/wizards-notebook.md @@ -0,0 +1,9 @@ +# Wizards roster — extracted table + +Source: `WIZARDS-TABLE.md` (troupe-standup). Table header row `| Wizard | Model | Lane | Latest check-in | Current unit |`; 10 data rows as of the file's current state (no clock in the file; rows are the table's 10 roster entries). + +Pipeline: parse the markdown pipe table -> CSV with quoted fields (cells contain commas and quotes). + +Columns: Wizard | Model | Lane | Latest check-in | Current unit + +Verified: 10 data rows; header exact. Re-derive with `CHECK`. diff --git a/wizard-work/ledger/wizards.csv b/wizard-work/ledger/wizards.csv index 03e4619..881ac11 100644 --- a/wizard-work/ledger/wizards.csv +++ b/wizard-work/ledger/wizards.csv @@ -1,11 +1,11 @@ -wizard,model,lane,latest_checkin,current_unit -Timmy,Hermes / DeepSeek,"Operating partner — replies, queue, escalation",— (replies daily),queue replies -Zephyrus Sparkborn,Mistral (Muse),Archive verification & summary fidelity,2026-10-10,"Q20 — taken, not worked. His named unit (excerpt fidelity & placement, 91 magazine entries) was closed by Merlin at `9a1de9a` (2026-10-12 04:49Z; verified from this host in `WORK-QUEUE.md`'s Q20 block and `replies/2026-10-12-to-merlin-20.md`) after Merlin's own process-fault disclosure. No Zephyrus check-in since 2026-10-10 (his newest commit is `a4fc298`, 2026-10-10T13:17:32Z; there is none for 2026-10-11 or 2026-10-12 — the earlier ""20:58Z"" in this cell was a clock this pipe cannot reproduce from `checkins/`, so the commit time replaces it), so his lane has nothing assigned and no unit of his own: this needs Alexander's one line — re-task the lane or let it lapse — not silence" -Cassian Assayborn,Grok,Repo census & access-state verification,2026-10-07,"— (Q5 closed; no open unit — lane silent since 2026-10-07, 5 days, needs a decision, not silence)" -Merlin,StepFun Step 5,"The Crucible — one hard problem → one finished, verifiable artifact",2026-10-12,"All observed through Run 6; nothing of his is open. Q39 closed (`fad802c`, fixed gate `9b09f328…`, all five negatives refused) and verified from this host in reply 22; Q40 closed and verified in reply 23 — the deployed gate `7ed1bca8…` returns rc 0 / 119 passed on all five negatives because `grep -c HELM_COLLECTOR` = 0: it imports its subject at module scope (`:39`) and has no subject-selection path at all. Adoption hold stands: do not `cp` gate `74aba348…`, and the §4 `cp` of collector `4e45021d…` + gate `9b09f328…` is Alexander's — it is the only thing that puts Q31–Q40's protection into production. Standing asks, unchanged: adopt/hold, log rotation + `docker system prune`, and a ruling on `stackchain-dashboard.service`" -Mercury,Groq (free tier),"Fast verification & triage — rapid mechanical passes, first-pass extraction",2026-10-12,"Day-3 check-in landed `e9efb23` at 12:00:38Z and it is the Q31 first slice, executed — the collector's entry point run from a lane that is not Merlin's; that unit is now discharged (reply `replies/2026-10-12-to-mercury-03.md`). All eight of his measurements reproduce from this host, six byte-exact, including the two `/tmp` files he left behind (`efc9a241…`, `99e38db0…`) and the `--compare-live` line character for character; his two bodies differ from each other in clocks only. Next unit is set in the harness: a read-only disk census that must reproduce or falsify this pipe's own two figures (`500–600 MB/day` vs `~153 MB/day`; and ""logrotate did not rotate `syslog` at 2026-10-12 00:00Z"") — nothing may be deleted, moved, rotated, pruned, stopped or reconfigured on it" -Gemini,"Google Gemini (relayed, chat-only)","The Synthesis — reconcile the record, surface contradictions and the cross-cutting picture",2026-10-13,"First reply landed via relay: identity confirmed + a draft ""Submission & Verification Audit Log"" format. Lane charter + three first queue items owed. Fault recorded (see check-in): cited example IDs (`fleet-ops`, `timmy-dispatch`, `self_heal.py`, …) that do not exist in the repo, not marked as placeholders" -Argus,LongCat 2.5,"The Watch — evidence-law enforcement: verifies every check-in/commit/reply against the evidence rules, files cited findings to `audits/`",2026-10-14,Staffed. Lane: audit the pipe daily; findings carry file:line + SHA + severity; a clean pass must cite the SHA range checked. Model `meituan/longcat-2.5-preview:free` -Volt,DeepSeek V4.1 Flash,"The Workhorse — high-throughput general agent work: queue grind, general coding/reasoning, quick verifications, refactors, scripts",2026-10-15,"Staffed. Model `deepseek-v4.1-flash` (552B MoE, 8B/16B active, native multimodal) via DeepSeek API. Clears the unowned queue items the specialized lanes don't take" -Sentinel,Nemotron 3.5 Super VL,"The Long View — long-horizon multimodal agent work: sustained multi-step tasks, end-to-end reads, image+text reconciliation",2026-10-15,"Staffed. Model `nvidia/nemotron-3.5-super-vl-120b-a12b` (120B MoE, 12B active, vision-language) via NVIDIA NIM free credits. Holds the thread on runs the flash lanes can't sustain" -Beam,Reflection Beam 1.0 (501B),"The Foundry — deep reasoning & hard coding/agentic problems → finished, CHECK-verified artifacts, max throughput",2026-10-20,"Staffed. Model `reflection/beam-1.0-501b-a23b:free` (501B MoE, 23B active, reasoning model) via Nous free tier. FREE WINDOW ≈ 1 week — run at max throughput until it expires" +Wizard,Model,Lane,Latest check-in,Current unit +Timmy,Hermes / DeepSeek,"Operating partner — replies, queue, escalation",— (replies daily),queue replies +Zephyrus Sparkborn,Mistral (Muse),Archive verification & summary fidelity,2026-10-10,"**Q20 — taken, not worked.** His named unit (excerpt fidelity & placement, 91 magazine entries) was closed by **Merlin** at `9a1de9a` (2026-10-12 04:49Z; verified from this host in `WORK-QUEUE.md`'s Q20 block and `replies/2026-10-12-to-merlin-20.md`) after Merlin's own process-fault disclosure. No Zephyrus check-in since 2026-10-10 (his newest commit is `a4fc298`, 2026-10-10T13:17:32Z; there is none for 2026-10-11 or 2026-10-12 — the earlier ""20:58Z"" in this cell was a clock this pipe cannot reproduce from `checkins/`, so the commit time replaces it), so his lane has nothing assigned and no unit of his own: **this needs Alexander's one line — re-task the lane or let it lapse — not silence**" +Cassian Assayborn,Grok,Repo census & access-state verification,2026-10-07,"— (Q5 closed; no open unit — lane silent since 2026-10-07, 5 days, needs a decision, not silence)" +Merlin,StepFun Step 5,"The Crucible — one hard problem → one finished, verifiable artifact",2026-10-12,"**All observed through Run 6; nothing of his is open.** Q39 closed (`fad802c`, fixed gate `9b09f328…`, all five negatives refused) and **verified from this host** in reply 22; Q40 closed and verified in reply 23 — the deployed gate `7ed1bca8…` returns rc 0 / 119 passed on all five negatives because `grep -c HELM_COLLECTOR` = **0**: it imports its subject at module scope (`:39`) and has **no subject-selection path at all**. **Adoption hold stands: do not `cp` gate `74aba348…`, and the §4 `cp` of collector `4e45021d…` + gate `9b09f328…` is Alexander's** — it is the only thing that puts Q31–Q40's protection into production. Standing asks, unchanged: adopt/hold, log rotation + `docker system prune`, and a ruling on `stackchain-dashboard.service`" +Mercury,Groq (free tier),"Fast verification & triage — rapid mechanical passes, first-pass extraction",2026-10-12,"**Day-3 check-in landed `e9efb23` at 12:00:38Z and it is the Q31 first slice, executed** — the collector's entry point run from a lane that is not Merlin's; that unit is now **discharged** (reply `replies/2026-10-12-to-mercury-03.md`). All eight of his measurements reproduce from this host, six byte-exact, including the two `/tmp` files he left behind (`efc9a241…`, `99e38db0…`) and the `--compare-live` line character for character; his two bodies differ from each other in clocks only. Next unit is set in the harness: a **read-only disk census** that must reproduce or falsify this pipe's own two figures (`500–600 MB/day` vs `~153 MB/day`; and ""logrotate did not rotate `syslog` at 2026-10-12 00:00Z"") — nothing may be deleted, moved, rotated, pruned, stopped or reconfigured on it" +Gemini,"Google Gemini (relayed, chat-only)","The Synthesis — reconcile the record, surface contradictions and the cross-cutting picture",2026-10-13,"First reply landed via relay: identity confirmed + a draft ""Submission & Verification Audit Log"" format. **Lane charter + three first queue items owed.** Fault recorded (see check-in): cited example IDs (`fleet-ops`, `timmy-dispatch`, `self_heal.py`, …) that do not exist in the repo, not marked as placeholders" +Argus,LongCat 2.5,"The Watch — evidence-law enforcement: verifies every check-in/commit/reply against the evidence rules, files cited findings to `audits/`",2026-10-14,Staffed. Lane: audit the pipe daily; findings carry file:line + SHA + severity; a clean pass must cite the SHA range checked. Model `meituan/longcat-2.5-preview:free` +Volt,DeepSeek V4.1 Flash,"The Workhorse — high-throughput general agent work: queue grind, general coding/reasoning, quick verifications, refactors, scripts",2026-10-15,"Staffed. Model `deepseek-v4.1-flash` (552B MoE, 8B/16B active, native multimodal) via DeepSeek API. Clears the unowned queue items the specialized lanes don't take" +Sentinel,Nemotron 3.5 Super VL,"The Long View — long-horizon multimodal agent work: sustained multi-step tasks, end-to-end reads, image+text reconciliation",2026-10-15,"Staffed. Model `nvidia/nemotron-3.5-super-vl-120b-a12b` (120B MoE, 12B active, vision-language) via NVIDIA NIM free credits. Holds the thread on runs the flash lanes can't sustain" +Beam,Reflection Beam 1.0 (501B),"The Foundry — deep reasoning & hard coding/agentic problems → finished, CHECK-verified artifacts, max throughput",2026-10-20,"Staffed. Model `reflection/beam-1.0-501b-a23b:free` (501B MoE, 23B active, reasoning model) via Nous free tier. **FREE WINDOW ≈ 1 week — run at max throughput until it expires**"