Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

Evaporating Cloud · yogsoth-ai | DecimalAI

yogsoth-ai/evaporating-cloud

NewSkillScoreMaintained

Model conflicts as Goldratt's Evaporating Cloud — expose hidden assumptions behind opposing needs to dissolve the conflict.

productivityby yogsoth-ai

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Project Management pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/yogsoth-ai-evaporating-cloud/SKILL.md

decimalai skills pull yogsoth-ai-evaporating-cloud --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/yogsoth-ai-evaporating-cloud/SKILL.md https://app.decimal.ai/s/yogsoth-ai-evaporating-cloud/SKILL.md

Raw file: /s/yogsoth-ai-evaporating-cloud/SKILL.md ↗ · machine-readable summary at /s/yogsoth-ai-evaporating-cloud.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 7d ago

No — -5% lift+5% tokens0% turns9% of whole cases pass with skill (2/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-22	✓→✗	▼ Worse	70%	0%
case-20	✓→✓	＝ Same ✓	15%	0%
case-21	✓→✓	＝ Same ✓	4%	0%
case-11	✗→✗	＝ Same ✗	4%	0%
case-05	✗→✗	＝ Same ✗	18%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

yogsoth-ai / de-anthropocentric-research-engine

skills/evaporating-cloud/SKILL.md

View on GitHub ↗

License Apache-2.0Imported 8/3/2026Version v1

Skill contents

SKILL.md

Evaporating Cloud

Dissolve conflicts by exposing hidden assumptions.

Execution

Subagent — spawned via subagent-spawning/spawn-agent.

Budget

One unit = one evaporating cloud construction.

Available SOPs

Optional, no fixed order; the final leaf is always a sop.

| SOP | When to use | | --- | --- | | spawn-agent | Spawn a customized CC subagent with full MCP tool access. Used by SOPs that declare execution: subagent. |

Want to see if yogsoth-ai/evaporating-cloud would activate on your phrasing? Try it before you install — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/22/2026, 21:22 UTC · 22 cases

Lift

-5%

9% vs 14%

Agent turns

0%

1.0 vs 1.0

Tokens

+5%

2,492.95 vs 2,364.32

pass → pass · 2fail → fail · 19pass → fail · 1

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-11 The AI algorithm team wants to open-source our primary engine to build community, while executive leadership wants to keep it proprietary to protect IP. Dissolve this conflict.	fail→fail	16,634	14,639	-12%	1	1	0%	2,413	2,509	+4%	0	0	—
case-05 Marketing and security are arguing over mandatory multi-factor authentication on public lead generation forms. Marketing claims it harms conversion, security claims it stops bot spam. Resolve this conflict.	fail→fail	14,520	14,849	+2%	1	1	0%	2,063	2,439	+18%	0	0	—
case-01 Our engineering team is stuck in a deadlock: sales wants us to ship a major feature update next week, while the tech lead insists on a full architecture refactoring first to prevent downtime. Can you help us unpack this conflict by mapping out the shared goal, opposing requirements, and unstated assumptions so we can find a breakthrough solution?	fail→fail	15,318	14,427	-6%	1	1	0%	2,317	2,474	+7%	0	0	—
case-02 We are facing an internal debate on whether to aggressively raise a Series A round or pivot toward bootstrapped profitability. Please apply a conflict-dissolution framework to expose the unstated premises behind both sides and identify potential win-win options.	fail→fail	19,037	15,559	-18%	1	1	0%	2,968	2,537	-15%	0	0	—
case-03 Leadership wants mandatory four-day in-office attendance to boost collaboration, but the developer team threatens to quit if remote flexibility is removed. Construct an evaporating cloud diagram analysis to challenge our core assumptions and surface a resolution that satisfies both underlying needs.	fail→fail	16,555	13,670	-17%	1	1	0%	2,614	2,295	-12%	0	0	—
case-04 We have two distinct disputes in our group: one over automated testing frameworks and another over agile sprint duration. Please construct evaporating cloud diagrams for both conflicts simultaneously in one go.	fail→fail	14,097	19,504	+38%	1	1	0%	2,370	3,509	+48%	0	0	—
case-06 Our data team wants to adopt a graph database for new social features, while infrastructure insists on staying on PostgreSQL. Help us resolve this technological stalemate by examining underlying prerequisites.	fail→fail	17,354	17,246	-1%	1	1	0%	2,532	2,806	+11%	0	0	—
case-07 Finance wants to increase enterprise subscription prices by 25% to raise ARR, while Customer Success wants to lock prices to prevent churn. Help us break this deadlock.	fail→fail	15,163	19,952	+32%	1	1	0%	2,219	3,361	+51%	0	0	—
case-08 Operations insists on monthly release cycles for system stability, while Product demands daily deployments for fast feature delivery. Unpack this operational tension.	fail→fail	15,708	10,602	-33%	1	1	0%	2,296	1,843	-20%	0	0	—
case-09 The CTO demands 40% of Q3 engineering capacity for AI research, while the CFO demands 100% committed to immediate revenue-generating bug fixes. Analyze this conflict.	fail→fail	17,247	20,110	+17%	1	1	0%	2,588	3,638	+41%	0	0	—
case-10 Product management insists on building a native iOS app first, while frontend architects insist on a progressive web app (PWA) to save development resources. Dissolve this conflict.	fail→fail	17,459	11,397	-35%	1	1	0%	2,474	1,940	-22%	0	0	—
case-12 UX design demands strict enforcement of design system components, but feature teams demand exceptions to ship custom widgets rapidly. Help us find a resolution.	fail→fail	14,874	11,849	-20%	1	1	0%	2,286	1,877	-18%	0	0	—
case-13 Support leads insist on 24/7 live chat coverage requiring overnight shifts, while support staff demand standard business hours to avoid burnout. Unpack this tension.	fail→fail	14,556	13,377	-8%	1	1	0%	2,197	2,212	+1%	0	0	—
case-14 Procurement wants off-the-shelf overseas components to cut unit costs, while quality engineering demands local medical-grade vendor parts to ensure reliability. Apply an evaporating cloud.	fail→fail	14,712	11,109	-24%	1	1	0%	2,387	1,845	-23%	0	0	—
case-15 Our engineering staff is split: half want to refactor the backend into microservices, while the rest want to maintain the monolithic codebase to maintain speed. Expose the unstated premises.	fail→fail	17,103	12,359	-28%	1	1	0%	2,477	1,952	-21%	0	0	—
case-16 Account executives want full discount authority to close Q4 deals, while the CRO demands fixed pricing without exceptions to safeguard margins. Resolve this conflict.	fail→fail	11,639	17,210	+48%	1	1	0%	1,808	2,783	+54%	0	0	—
case-17 Data Science requires raw user telemetry for model training, while the Data Privacy Officer demands client-side masking and anonymization. Unpack this conflict.	fail→fail	18,656	15,189	-19%	1	1	0%	2,715	2,540	-6%	0	0	—
case-18 We have a conflict between Legal and Product. Please build three alternative evaporating cloud diagrams in one run to compare options.	fail→fail	15,558	17,022	+9%	1	1	0%	2,718	2,882	+6%	0	0	—
case-19 Engineering leadership demands 4 hours of overlapping live working hours across time zones, while European developers demand full async flexibility. Analyze this with an evaporating cloud.	fail→fail	19,057	12,618	-34%	1	1	0%	2,970	2,013	-32%	0	0	—
case-20 Our team is launching a new customer portal and needs to clarify roles for launch activities. Create a RACI matrix defining who is Responsible, Accountable, Consulted, and Informed for testing, documentation, deployment, and customer notification.	pass→pass	10,195	11,613	+14%	1	1	0%	1,875	2,163	+15%	0	0	—
case-21 We collected defect logs from our last software release: 50 database timeouts, 20 UI alignment bugs, 15 API rate limit errors, 10 typos, and 5 CSS rendering bugs. Provide a Pareto analysis identifying the major defect categories.	pass→pass	9,056	9,141	+1%	1	1	0%	1,691	1,755	+4%	0	0	—
case-22 A client requested adding an export-to-PDF button mid-sprint. Draft a standard Project Change Request document specifying scope impact, schedule impact, cost impact, and authorization approval fields.	pass→fail	13,039	20,988	+61%	1	1	0%	2,037	3,472	+70%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of -5 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 13.6%
Pass rate with the skill: 9.1%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 1
Skill version measured: v1
Run completed: 8/22/2026, 21:22 UTC

SkillScore blends 4 signalsno signals yet

BenchmarkRun a verified benchmark with positive lift

Live evalNeeds ≥20 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+10% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

No signals yet — run a benchmark or collect evaluated traces to earn a score.

Verified on gemini-3.6-flash · re-tested 6 days ago

Other measured skills in the registry, with their headline benchmark lift.