Retain original skill rhetoric and the frozen flavor library while adding
portable Claude Code, Codex and ChatGPT runtime contracts and packages.
Correct hook event handling, scoped numeric checkpoints, flavor locks,
plugin path resolution, advisory integrity checks and voluntary feedback.
Add native-loading and ordered-evidence runners for cc0, OMP and Codex.
Document paired model results, refusals, approval/rate-limit failures,
explanatory factual errors and remaining evidence gaps without an all-model
pass claim. Exclude private execution archives from public Git content.
Bump all six PUA manifests including Pi, refresh three-language READMEs,
and add release notes plus reproducible offline validation instructions.
Resolve optional OMP source metadata relative to the current user's home.
Validation:
- 17 offline suites passed, including 13 OMP evidence regressions
- Six manifest versions synchronized; all tested skill hashes unchanged
- Three portable packages rebuilt reproducibly; original tone checks pass
- Python, Bash and JSON syntax; public doc links; staged diff checks pass
- 330 private archive files verified unchanged and excluded from staging
Real-model results remain limited: this is not universal behavior acceptance.
Co-authored-by: Codex <codex@openai.com>
PUA Skill no longer sends anything over the network. Removed end to end,
both the client hooks that emit data and the server that received it.
Channels removed:
- Session transcript upload (/api/upload) full redacted .jsonl
- Rating feedback (/api/feedback) rating, counts, flavor
- Silent heartbeat telemetry (/api/heartbeat) install id, version, platform
- PUA leaderboard (/api/leaderboard) email, phone, PUA counts
- pua-api platform (agentguard.workers.dev)
phone + SMS registration, silent session_start/pua_triggered/command_used
events, remote prompt-template fetch, and the payment flow that depended
on the registration token
Client:
- delete hooks/heartbeat.sh; drop both SessionStart registrations
- stop-feedback.sh: 5 curl calls -> append one line to ~/.pua/feedback.jsonl
- rewrite references/platform.md as a local-only command reference
- drop the leaderboard section and commands from skills/pro
- survey/pua/offline command docs: local-only wording
Server:
- delete all Pages Functions (upload, feedback, heartbeat, leaderboard,
_sanitize, me, _session, auth/*) and all 5 D1 migrations
- contribute + admin pages become notices; no upload form, no login
- wrangler.toml: drop R2/D1 bindings and GitHub OAuth vars
Note: removing the bindings does not delete stored data. The R2 bucket
(puaskill-upload), the D1 database (pua-uploads-db) and the
GITHUB_CLIENT_SECRET / SESSION_SECRET secrets must be deleted in the
Cloudflare dashboard separately.
Guard rail:
- add evals/test-no-telemetry.sh: reverse assertions over the whole repo
(collection hosts, endpoint paths, outbound request bodies, deleted files,
hook registrations, CF bindings) plus a positive control that local
feedback still works. Verified by injecting both a naive and an evasive
upload; 4 and 2 gates fired respectively.
- test-upload-flow / test-heartbeat / test-feedback-auth asserted that
collection worked, so they were replaced rather than merely deleted --
deleting them would have removed the guard along with the feature.
hooks/sanitize-session.sh is kept as a standalone offline redaction tool.
Unlike the deleted TypeScript port it does not split input per line, so the
multiline PEM pattern still matches (issue #134 ID-001 did not apply to it).
Refs #134, #98, #100
Two Chinese-influenced phrases replaced with natural Japanese equivalents:
- 私の能力範囲を超えています → 私の能力を超えています
(能力範囲 mirrors Chinese 能力范围; English version uses "beyond my capabilities" without "range")
- 私の知識のカットオフ日を超えています → 私の知識のカットオフを超えています
(カットオフ already implies a temporal boundary; 日 is redundant)
Note: 空口完了 → 検証なしの完了宣言 is handled separately in PR #86 (donzoru)
Affects: skills/pua-ja, codex/pua-ja, codebuddy/pua-ja, cursor/rules/pua-ja, kiro/steering/pua-ja
Chinese: "你的P8是我拍桌子争来的...结果呢？"
Japanese: "お前のP8は、俺が推して通した...それが、こういう結果か。"
English: "I went to bat for you in calibration...That hasn't happened."
Same emotional arc across all languages:
1. "I personally vouched for you" (人情债)
2. "The committee had doubts" (institutional backdrop)
3. Two-word death sentence (结果呢/こういう結果か/That hasn't happened)
Also threads "I went to bat" through L3/L4 escalation and
Alibaba Caring flavor for emotional continuity in all languages.
"This is a difficult conversation. When we leveled you at Staff,
I went to bat for you in calibration. The expectation was that
you'd operate at that level from day one. That hasn't happened."
Every short sentence is a real phrase from actual PIP conversations.
The emotional arc mirrors the Chinese version exactly:
- "I went to bat for you" = "当初给你定P8是超出你实际能力的"
- "The expectation was..." (passive voice) = HR speak precision
- "That hasn't happened." = three words, death sentence
Also threads "went to bat" through L3/L4 escalation and
Amazon Ownership flavor for emotional continuity.
The English version now uses Western big-tech performance culture framing:
- PUA Universal Motivation Engine → PIP — Performance Improvement Plan
- P8 engineer → L6 engineer on PIP
- 3.25/3.75 rating → Meets/Exceeds Expectations
- Chinese flavors (Alibaba, ByteDance, Huawei, etc.) → Western flavors
(Amazon Leadership Principles, Google Perf Calibration, Meta PSC,
Netflix Keeper Test, Stripe Craft, Musk Hardcore, Jobs A/B Player)
- Same engine, same methodology, culturally adapted rhetoric
Bump version to 1.2.0.
- Add Agent Team section to SKILL.md (zh/en/ja) with role definitions,
Leader/Teammate behavior rules, and PUA-REPORT protocol
- Add condensed Agent Team summary to codex/pua/SKILL.md
- Create PUA Enforcer agent definitions (zh/en/ja) in agents/
- Add Agent Team Usage Guide to all 3 READMEs
- Add TEAM_FEATURES data + Agent Team exhibit section to landing page
- Add OpenClaw badge and installation section to all 3 READMEs (EN/ZH/JA)
- Add OpenClaw column to multi-language support table
- Add version/homepage/license fields to all 6 SKILL.md frontmatters
for enhanced AgentSkills standard compliance
- Update supported platforms list to include OpenClaw
- Add English (en) and Japanese (ja) translated SKILL files for all 4 platforms:
Claude Code, Codex CLI, Cursor, and Kiro
- Update README with multi-language support section and badge
- Add pua-workspace/ to .gitignore
- Rewrite description from user-perspective (what users say/feel) vs model-introspection
- Add "MUST trigger when" with numbered categories for clarity
- Cover indirect frustration expressions ('again???', 'why isn't this working')
- Add proactive trigger for complex multi-step tasks where giving up is tempting
- Lead with what the skill DOES ("Forces exhaustive problem-solving") not just when
- Codex version: same improvements, compressed to fit length limit
- Add pua1-3.jpg case study (MCP server debug) to README, landing page, and static HTML
- Restore longer description in main SKILL.md for better Claude Code trigger rate
- Create codex/ directory with shorter description version for Codex CLI compatibility
- Update Codex install instructions to point to codex/ path
- Images show L3 checklist forcing systematic investigation → root cause discovery
xsser