test(agent_patterns): S5 scenario diversification — categories beyond coding + C16 spread#80
Merged
Merged
Conversation
added 2 commits
June 6, 2026 20:22
… coding + C16 spread Roadmap S5 (2026-06-05). Catalog-only change (no harness code), built on S1's live:optional + multi-op groundwork. New patterns (6, all live PASS individually and in the full run): - repeat_visits x2 in workflows.yaml (stateful multi-op stays in one file so the catalog-unique-URL rule remains auditable): BBC /innovation news monitoring loop, CoinGecko ETH price polling loop. Both assert C8 cache_hit on revisit + visit-3 suggest_profile (threshold verified deterministic in pipeline.VISIT_HINT_THRESHOLD=3). - error_handling x2 in domain shards: dead yna.co.kr article URL (news), hallucinated Yahoo Finance ticker (finance). Graceful no-crash only. - large_page x2 in wiki_reference: en World_War_II (observed 7.2-9.0 s), ko 대한민국 CJK large article (observed 5.8-8.2 s); both carry the n_chunks_max=12 / output_chars_max=8000 budgets from the existing Korean_War pattern. C16 enrichment assertions spread to 5 existing single_fetch patterns, all non-coding so the coding 24/24 surface is untouched: - excerpts_min_count: hermes_hn_front_page, hermes_fred_cpi, claude_code_wiki_en_yi_sunsin - chain_hints_has_key: wiki en/ko yi_sunsin (search_template), ja sushi (recommended_followup_filter) - page_entities_contain_any: ko yi_sunsin (probe-observed 임진왜란/해전; en wiki entities observed empty — title not in heading_paths — so en asserts excerpts/chain_hints instead) - outbound_links_contain_any left unused: live probes found no stable surface (HN links rotate with stories, GitHub had 1 top-k-dependent link, wikipedia fetcher yields 0) Gate: full 110-pattern shared-state run 101 PASS + 9 SKIP (same live:optional set as before S5), exit 0. dry-run schema 110/110.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Roadmap S5 (시나리오 다양화) — the last open item from
docs/superpowers/plans/2026-06-05-trawl-improvement-roadmap.md. Catalog-only change (no harness code), as the S5 gate requires.New patterns (6)
hermes_bbc_tech_repeat_checkopenclaw_coingecko_eth_repeat_checkopenclaw_error_dead_news_articlehermes_error_yahoo_invalid_tickerhermes_wiki_en_world_war_ii_largeopenclaw_wiki_ko_korea_largeMulti-op repeat_visits patterns stay in workflows.yaml per the README convention so the stateful catalog-unique-URL rule (S1 lesson) remains auditable in one file; both use catalog-unique URLs (
/innovation,/coins/ethereum).C16 enrichment spread (5 existing single_fetch patterns, all non-coding)
excerpts_min_count: hermes_hn_front_page, hermes_fred_cpi, claude_code_wiki_en_yi_sunsinchain_hints_has_key: wiki en/ko yi_sunsin (search_template), ja sushi (recommended_followup_filter)page_entities_contain_any: ko yi_sunsin (probe-observed 임진왜란/해전)outbound_links_contain_anydeliberately unused — live probes found no drift-stable surface (HN links rotate, GitHub 1 top-k-dependent link, wikipedia fetcher 0 links)Gate evidence (pre-registered in roadmap S5)
--dry-run: 110/110 schema valid