clickhouse-autoresearch-campaign

Run a ClickHouse query optimization campaign on one git branch using pi-autoresearch, dynamic lanes and hypotheses, baseline result capture, correctness checks, and stagnation-aware lane/campaign review.

Quality

61%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./tools/query-performance-ai/query_performance_ai/sandboxed_autoresearch_agent/pi_plugin/skills/clickhouse-autoresearch-campaign/SKILL.md

Quality

Content

63%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with concrete commands and strong validation/feedback loops, but it is verbose and inlines reference-grade material that should live in a separate file, and its key sibling references are not bundled. Splitting the profiling reference material into a bundled file and fixing the duplicated step numbering would lift the weaker dimensions.

Suggestions

Move the 'Adapter capabilities' EXPLAIN-variant, system.* table, and ProfileEvents reference material into a bundled reference file (e.g. references/profiling.md) and link to it one level deep.

Ensure the referenced `orchestration.md` and the `ch_*.py` scripts are actually bundled under references/ and scripts/ so signaled references resolve.

Fix the Setup sequence numbering: step 6 appears twice (capture baseline, then read baseline) with a nested 1–5 list in between; renumber to a single clean 1–N sequence.

Dimension	Reasoning	Score
Conciseness	It mostly assumes Claude's competence and avoids explaining basic concepts, but the ~260-line body inlines a large reference block (every EXPLAIN variant, four system.* profiling tables, ProfileEvents glossary, source-browsing tips) that could be tightened or moved to a reference file.	3 / 5
Actionability	Provides concrete, executable bash and SQL throughout — `ch_campaign_init.py` with enumerated flags, `ch_capture_baseline.py`, EXPLAIN and system.* lookup queries — but some rely on placeholders (`<query_id>`, `<id>`) and sibling scripts whose availability is assumed, leaving minor gaps.	4 / 5
Workflow Clarity	The Setup sequence and Runtime responsibilities are clearly ordered with real validation checkpoints (baseline capture, `autoresearch_checks.py` correctness backpressure, escalation checks), but the Setup sequence contains a duplicate 'step 6' and a nested numbered list that muddies the ordering.	4 / 5
Progressive Disclosure	Section headers give it structure and it signals an `orchestration.md` sibling, but no bundle files (references/scripts/assets) are present, the referenced orchestration contract and scripts are not bundled, and large reference-style material (EXPLAIN/system.* tables) is inlined rather than split into a one-level-deep reference file.	3 / 5
	Total	14 / 20 Passed

Description

58%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific and distinctive but over-relies on internal orchestration terminology and omits any explicit 'Use when...' trigger guidance, capping completeness. Adding natural user-facing trigger phrases would materially raise trigger_term_quality and completeness.

Suggestions

Add a 'Use when ...' clause naming natural user triggers such as 'Use when optimizing a slow ClickHouse query, reducing ClickHouse query latency, or running an automated query-tuning campaign.'

Replace internal jargon ('dynamic lanes', 'stagnation-aware review') in the description with user-recognizable terms, or move that detail into the body.

Broaden trigger-term coverage with synonyms like 'ClickHouse query tuning', 'slow query', and 'query performance' so users' natural phrasing matches.

Dimension	Reasoning	Score
Specificity	Names the domain (ClickHouse query optimization) plus several concrete actions — 'baseline result capture', 'correctness checks', 'stagnation-aware lane/campaign review', 'dynamic lanes and hypotheses' — but the actions are process-oriented sub-steps rather than a comprehensive set of distinct capabilities, leaving minor coverage gaps.	4 / 5
Completeness	It clearly states 'what' (run a ClickHouse query optimization campaign) but provides no 'Use when...' clause or equivalent trigger guidance, so per the rubric completeness is capped at 3.	3 / 5
Trigger Term Quality	Contains one strong natural phrase ('ClickHouse query optimization') but is otherwise dominated by internal jargon ('pi-autoresearch', 'dynamic lanes', 'stagnation-aware', 'hypotheses') that users would rarely say verbatim, missing common variations and synonyms.	3 / 5
Distinctiveness Conflict Risk	The 'ClickHouse query optimization campaign on one git branch using pi-autoresearch' niche is clearly distinct with specific triggers, with only minor overlap risk against general query-optimization skills.	4 / 5
	Total	14 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: PostHog/posthog
Path: tools/query-performance-ai/query_performance_ai/sandboxed_autoresearch_agent/pi_plugin/skills/clickhouse-autoresearch-campaign/SKILL.md
Commit: 6fca5f8

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.