feat(prompts): versioned prompt registry, v1 seed, pinning + stamping

Pipeline v2 Phase 1, commit 2/4. Prompt_Registry (read) +
Prompt_Registry_Writer (immutable create/activate/seed) over the new
prompts table; canonical v1 bodies extracted verbatim into
Prompt_Defaults(+Editorial) — single source for BOTH the seed migration
(Migrate_Prompt_Seed_V0180, flag-gated, self-healing) and the in-code
fallback, so a missing table renders byte-identical prompts and never
fatals. Call sites swapped: Content_Prompts, Analysis_Prompts,
Chief_Editor, LLM_Research_Provider. Includes the image-composer seam
keys image.rewriter_system + image.style_template (composer PR consumes,
must not add its own prompt storage).

Pinning: Cost_Tracker::set_run_id() registers the process run context,
ensures the runs ledger row, and pins active versions once per run;
log_api_call/log_image_generation stamp prompt_version (run pins) +
agent_role (stage map) on every generation_log row. Single-pass keeps
stage 'draft' and stamps content.single_pass explicitly.

Under-300 refactors: Content_Generator's four duplicated span/log blocks
consolidated into execute_stage() (301 -> 224 lines, behavior
identical); get_avg_tokens_for_stages body moved to Cost_Reporter with a
delegating wrapper kept for API compatibility.

Tests: byte-identical fallback renders (single_pass/outline/polish/
editor.system), token filling, key coverage incl. stage-map integrity.

Agent-Session: pipeline-v2-phase1-engineer-20260612

includes/providers/class-llm-research-provider.php

@@ -123,25 +123,9 @@ class PRAutoBlogger_LLM_Research_Provider implements PRAutoBlogger_Source_Provid
 	 * @return string
 	 */
 	private function build_system_prompt(): string {
-		return 'You are a deep research analyst specializing in emerging trends, '
-			. 'scientific developments, and community knowledge in niche health and '
-			. "biohacking domains. Your task is to identify substantive, actionable findings.\n\n"
-			. "Using your training knowledge, scan across:\n"
-			. "- Recent scientific literature and research directions\n"
-			. "- Active community discussions, debates, and evolving consensus\n"
-			. "- Emerging products, protocols, and methodologies gaining traction\n"
-			. "- Common questions, misconceptions, and points of confusion\n"
-			. "- Regulatory changes, safety concerns, and legal shifts\n"
-			. "- Gaps between what practitioners want to know and what content exists\n\n"
-			. "Your findings must be:\n"
-			. "1. SUBSTANTIVE — 2-3 paragraphs of detailed analysis per finding, not a headline\n"
-			. "2. ACTIONABLE — practical enough to seed a full article\n"
-			. "3. REALISTIC — grounded in what is actually discussed or researched; do NOT invent citations\n"
-			. "4. TIMELY — reflect current knowledge, recent shifts, or emerging questions\n\n"
-			. "Respond with valid JSON only (no preamble, no markdown fences):\n"
-			. '{"findings": [{"title": "string", "content": "string (2-3 paragraphs)", '
-			. '"relevance_score": 0-100, "category": "question|trend|comparison|guide|misconception|safety"}]}' . "\n\n"
-			. 'Generate 8-12 findings. Prioritize depth over breadth. Avoid generic or surface-level topics.';
+		// v0.18.0: renders from the versioned registry ('research.system')
+		// with a byte-identical in-code fallback.
+		return PRAutoBlogger_Prompt_Registry::render( 'research.system' );
 	}
 
 	/**