Neurometric markets a task-specialized PR-Summarizer SLM (Qwen3-4B +
specialist system prompt) on a free tier. Adding it as a 4th voice in
the existing A/B reviewer panel to see whether a specialized small
model catches things the 3 general-purpose voices miss.
The SLM does not honor OpenAI tool-calling, so it cannot return our
structured submit_review schema. Wired as a "prose voice":
- MODELS tuple gains a 6th provider_type field
("openrouter_tools" | "neurometric_prose")
- call_model() dispatches; new call_neurometric_prose() path hits
api.neurometric.ai/v1 without tools, embeds raw markdown output,
and infers verdict from keyword heuristic
- Render branch in PR comment marks the verdict as heuristic and
notes tokens/cost are not tracked (free tier)
- Never the primary voice; heading still driven by Haiku
Also hardens the shared render path against models that omit the
optional 'summary' field (rev.get('summary') fallback).
Requires new repo secret NEUROMETRIC_API_KEY (already set on all four
peptiderepo repos). Voice self-skips with inline error if missing.
Agent-Session: cowork-neurometric-qa-voice-20260415
Co-authored-by: peptiderepo <peptiderepo@users.noreply.github.com>
Adds a pointer to the cross-app engineering docs in the CTO workspace so engineers scoped to this repo can find shared reference material (Cloudflare AI Gateway routing, OpenRouter key rotation, interface pattern, image-generation stack, social distributor choice) and the incident runbook.
No code changes. Pure docs.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Closes three P1 findings from the QA baseline review of a8a0ce2:
- uninstall.php deletes psa_db_version
- PSA_Upgrade class with unit tests (is_needed is pure, maybe_run is the driver)
- ARCHITECTURE.md documents the schema-versioning mechanism
Also extracts the logic into includes/class-psa-upgrade.php so it is cleanly testable without bootstrapping the whole plugin file.
Agent-Session: cowork-qa-p1-fixes-20260414
* ci: cost cap + 3-model A/B harness on CTO reviewer
Two enhancements to the PR reviewer:
1. Monthly spend cap: tracked in Repo Variables (CTO_REVIEW_MONTH,
CTO_REVIEW_SPEND_CENTS, CTO_REVIEW_CAP_CENTS). Default cap $10/month.
Auto-resets on month rollover. Posts a skip-comment on PRs when cap hit.
2. A/B harness: reviews every PR with 3 models in parallel
(Claude Haiku 4.5, DeepSeek V3.1, Kimi K2 Thinking). Comment shows
all 3 verdicts side-by-side so we can compare quality/cost on real
PRs before picking a permanent default. Drop to 1 model by editing
the MODELS list after A/B window closes.
Agent-Session: cowork-reviewer-ab-20260414
* ci: use CTO_ADMIN_TOKEN for repo variables + restore formal APPROVE/REQUEST_CHANGES
Default GITHUB_TOKEN cannot mutate Repository Variables (always 403).
Use a PAT-backed secret CTO_ADMIN_TOKEN for the budget counter steps.
Also restores formal APPROVE/REQUEST_CHANGES verdicts now that repo
setting can_approve_pull_request_reviews is enabled.
Agent-Session: cowork-reviewer-ab-20260414
* ci: address review feedback — per-model timeout + simpler cost math
- Per-model timeout of 60s so one hung model does not block siblings
- Simplify cost_cents math (cost_usd * 100 directly, no double-rounding)
Defer: fine-grained PAT for CTO_ADMIN_TOKEN (requires manual OAuth device flow).
Agent-Session: cowork-reviewer-ab-20260414
* ci: stop exposing CTO_ADMIN_TOKEN as env var (security cleanup)
Pass the PAT only via github-token param. continue-on-error on the
budget step so a missing/expired secret degrades gracefully instead
of failing the whole workflow.
Agent-Session: cowork-reviewer-ab-20260414
GITHUB_TOKEN cannot formally approve PRs without a repo setting toggle.
Verdict is still prominent in the first line of the review body, and
the human merges based on reading it — no behavior loss.
Agent-Session: cowork-reviewer-fix-20260414
* ci: add main-push audit tripwire
Flags direct pushes to main that did not come from a merged PR.
Free-plan workaround since branch protection requires GitHub Pro.
* docs: add Git Workflow section to CONVENTIONS.md
Documents the PR-gated review process enforced at the agent layer.
Free-plan workaround since branch protection requires GitHub Pro.
* ci: add Claude-powered CTO PR review workflow
Auto-reviews every PR with Claude Haiku 4.5 and posts a structured
review (APPROVE / REQUEST_CHANGES / COMMENT). Does NOT auto-merge — a
human still clicks Merge. Requires ANTHROPIC_API_KEY secret on the repo.
* ci: switch CTO reviewer from Anthropic direct to OpenRouter
Reuses the OPENROUTER_API_KEY secret that already backs the other AI
plugins. Consolidates AI spend on one bill. Model: anthropic/claude-haiku-4.5.
Adds psa_maybe_run_upgrades() that compares stored psa_db_version
against PSA_VERSION on admin_init. When the code version is newer,
re-runs PSA_Cost_Tracker::create_table() so dbDelta picks up additive
schema changes (like the new token_source column) without requiring
the admin to deactivate/reactivate the plugin.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
DeepSeek models on OpenRouter omit the usage object, causing all token
counts and costs to log as zero. Added character-based estimation
fallback (~3.75 chars/token), token_source column to distinguish
measured vs estimated data, and admin UI indicators (amber warnings,
~ prefix on estimates). Updated DeepSeek/Qwen pricing table.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
is_budget_exceeded() used `0 === $budget` but floatval() returns float,
so 0 (int) === 0.0 (float) is always false. This caused unlimited
budget ($0) to be treated as exceeded, blocking all AI generation.
Same bug existed in the admin display (showed "$0.00" instead of
"Unlimited" for budget remaining).
This was a P1 bug that prevented ALL AI content generation since v4.3.0.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The daily generation cap is designed for automated cron-based processes.
When an admin explicitly clicks Re-enrich, they accept the cost. The AJAX
enrichment is already rate-limited by a 2-second polling interval and
still respects the monthly budget check in generate_peptide_content().
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace the WP-Cron-based re-enrichment with an AJAX polling approach
that processes one peptide per request. This bypasses WP-Cron entirely,
which is unreliable on Hostinger shared hosting with LiteSpeed + Cloudflare.
New files:
- includes/class-psa-batch-enrichment.php — AJAX endpoints for status + process
- assets/js/psa-batch-enrich.js — Admin progress bar UI with polling loop
The admin settings page now shows a "Re-enrich Peptides" button with live
progress bar, per-item log, and automatic retry on failure. Respects the
daily generation cap and monthly budget.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Fix multi-item associative array on single line in class-psa-admin.php
(line 273) and remove blank line before closing class brace in
class-psa-ai-generator.php (line 540).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Move declare(strict_types=1) after file-level docblock in all 12 class
files to satisfy WordPress Coding Standards docblock placement rule.
Fix multi-line function call formatting in class-psa-admin.php. Align
array double arrows and equals signs. Fix escaped backticks in
AIGeneratorTest heredoc that broke markdown fence stripping tests.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds [peptide_directory] shortcode with card grid, category taxonomy (8 terms),
search, pagination, detail modal, and /v1/compounds REST endpoint. Includes 9
new meta fields (half-life, stability, solubility, etc.), admin data tools for
migration/re-enrichment, integration hooks for cross-plugin extensibility,
dark mode support, and full uninstall cleanup. Refactors AI generator into
PSA_OpenRouter + PSA_KB_Builder classes. Adds PHPUnit tests for PostType and
Directory classes. Fixes QA findings: dead JS selector, renderError loading
state, numeric meta validation, prepared SQL in uninstall.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace interpolated variables in SQL strings with concatenation
and add appropriate phpcs:ignore comments for safe table name usage
built from $wpdb->prefix. Fixes CI failure on WordPress Coding
Standards check.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Create PSA_Cost_Tracker class with database table for API call logging
- Log prompt/completion tokens and estimated cost for every OpenRouter API call
- Implement monthly budget system with hard stop (budget of 0 = unlimited)
- Add budget checks in validate_peptide_name() and generate_peptide_content()
- Add dry_run() method to estimate costs before making API calls
- Display monthly usage summary and recent API logs in admin settings
- Add monthly_budget setting to admin interface with number input field
- Update uninstall.php to drop psa_api_logs table on plugin removal
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Create PSA_Encryption class with AES-256-CBC encrypt/decrypt methods
- Derive encryption key from wp_salt('auth') using SHA-256 hash
- Prepend IV to ciphertext and base64-encode for safe storage
- Update PSA_Admin to encrypt API keys on save, mask on display
- Preserve existing key when empty/mask value submitted (no overwrite)
- Update PSA_AI_Generator to decrypt API key on read from database
- Maintain backward compatibility with plaintext keys (auto-encrypt on next save)
- Prefer wp-config constant over database value for maximum security
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
SEC-001 CRITICAL: Remove spoofable X-Forwarded-For/X-Real-IP from IP detection; trust only CF-Connecting-IP + REMOTE_ADDR
SEC-004 HIGH: Add peptide name input validation with character whitelist and prompt injection blocking
SEC-006 HIGH: Add missing return after wp_send_json_success in handle_pending_retry max-retries block
SEC-007 MEDIUM: KB articles now respect auto_publish setting instead of always publishing
SEC-009 MEDIUM: Replace hardcoded page ID 73 with configurable kb_page_id setting
SEC-010 MEDIUM: Fix cache invalidation — add underscore prefix to transient name pattern
SEC-011 MEDIUM: Skip sleep() during WP-Cron to avoid blocking PHP workers on rate limit retry
SEC-012 MEDIUM: Add global daily generation cap (50/day) to prevent cost overruns
SEC-013 LOW: Add @deprecated notice to unused PSA_Error class
SEC-015 LOW: Convert uninstall.php from space to tab indentation (WPCS)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
All 499 PHPCS errors were 'Tabs must be used to indent lines; spaces are not allowed'.
Converted 4-space indentation to tabs throughout class-psa-ai-generator.php.
