{"id":"fedd9517-01df-4770-ae2d-668749abedda","shortId":"geqm4A","kind":"skill","title":"ralph","tagline":"Persistent completion mode. Use when the user explicitly says `/ralph` or clearly wants you to keep iterating until the task is actually finished, repeating implement-verify-fix loops instead of stopping at partial progress.","description":"# Ralph\n\nUse this as a persistence override, not a separate methodology.\n\n## Default\n\n- Treat the task as incomplete until the requested outcome is verified.\n- Repeat the loop: inspect -> implement -> verify -> fix -> verify again.\n- If a check fails, continue with the next fix instead of stopping at the first attempt.\n- If one approach stalls, switch approaches and keep going.\n- Keep intermediate updates short and progress-focused.\n\n## Do Not Stop For\n\n- \"Probably fixed\"\n- \"Good enough\"\n- Partial completion\n- A failed first or second attempt\n- Needing one more verification pass\n\n## Stop Only When\n\n- The requested outcome is verified.\n- A real external blocker prevents further progress.\n- The user explicitly stops or redirects the work.\n\n## Avoid\n\n- Do not ask whether to continue after each failure.\n- Do not declare success without running the relevant checks.\n- Do not fall back to planning mode unless execution is genuinely blocked.\n\n## Finish\n\nReturn:\n- what was completed\n- what was verified\n- what is still blocked, if anything","tags":["ralph","dgk","gpt","dgk-dev","agent-skills","cli","codex","developer-tools","mcp","openai-codex"],"capabilities":["skill","source-dgk-dev","skill-ralph","topic-agent-skills","topic-cli","topic-codex","topic-developer-tools","topic-mcp","topic-openai-codex"],"categories":["dgk-gpt"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/dgk-dev/dgk-gpt/ralph","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add dgk-dev/dgk-gpt","source_repo":"https://github.com/dgk-dev/dgk-gpt","install_from":"skills.sh"}},"qualityScore":"0.472","qualityRationale":"deterministic score 0.47 from registry signals: · indexed on github topic:agent-skills · 45 github stars · SKILL.md body (1,045 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-05-02T00:57:07.693Z","embedding":null,"createdAt":"2026-04-18T22:17:59.034Z","updatedAt":"2026-05-02T00:57:07.693Z","lastSeenAt":"2026-05-02T00:57:07.693Z","tsv":"'/ralph':11 'actual':23 'anyth':190 'approach':87,90 'ask':149 'attempt':84,117 'avoid':146 'back':168 'block':176,188 'blocker':134 'check':71,164 'clear':13 'complet':3,111,181 'continu':73,152 'declar':158 'default':48 'enough':109 'execut':173 'explicit':9,140 'extern':133 'fail':72,113 'failur':155 'fall':167 'finish':24,177 'first':83,114 'fix':29,66,77,107 'focus':101 'genuin':175 'go':93 'good':108 'implement':27,64 'implement-verify-fix':26 'incomplet':53 'inspect':63 'instead':31,78 'intermedi':95 'iter':18 'keep':17,92,94 'loop':30,62 'methodolog':47 'mode':4,171 'need':118 'next':76 'one':86,119 'outcom':57,128 'overrid':43 'partial':35,110 'pass':122 'persist':2,42 'plan':170 'prevent':135 'probabl':106 'progress':36,100,137 'progress-focus':99 'ralph':1,37 'real':132 'redirect':143 'relev':163 'repeat':25,60 'request':56,127 'return':178 'run':161 'say':10 'second':116 'separ':46 'short':97 'skill' 'skill-ralph' 'source-dgk-dev' 'stall':88 'still':187 'stop':33,80,104,123,141 'success':159 'switch':89 'task':21,51 'topic-agent-skills' 'topic-cli' 'topic-codex' 'topic-developer-tools' 'topic-mcp' 'topic-openai-codex' 'treat':49 'unless':172 'updat':96 'use':5,38 'user':8,139 'verif':121 'verifi':28,59,65,67,130,184 'want':14 'whether':150 'without':160 'work':145","prices":[{"id":"81463b5d-997d-485b-84dc-498db6d5e2fa","listingId":"fedd9517-01df-4770-ae2d-668749abedda","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"dgk-dev","category":"dgk-gpt","install_from":"skills.sh"},"createdAt":"2026-04-18T22:17:59.034Z"}],"sources":[{"listingId":"fedd9517-01df-4770-ae2d-668749abedda","source":"github","sourceId":"dgk-dev/dgk-gpt/ralph","sourceUrl":"https://github.com/dgk-dev/dgk-gpt/tree/main/skills/ralph","isPrimary":false,"firstSeenAt":"2026-04-18T22:17:59.034Z","lastSeenAt":"2026-05-02T00:57:07.693Z"}],"details":{"listingId":"fedd9517-01df-4770-ae2d-668749abedda","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"dgk-dev","slug":"ralph","github":{"repo":"dgk-dev/dgk-gpt","stars":45,"topics":["agent-skills","cli","codex","developer-tools","mcp","openai-codex"],"license":"mit","html_url":"https://github.com/dgk-dev/dgk-gpt","pushed_at":"2026-03-13T04:37:05Z","description":"Safe Codex CLI setup installer with skills, AGENTS.md, profiles, and MCP defaults","skill_md_sha":"c9ea534c8760af6b908ae143ab8a99cdc97f340c","skill_md_path":"skills/ralph/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/dgk-dev/dgk-gpt/tree/main/skills/ralph"},"layout":"multi","source":"github","category":"dgk-gpt","frontmatter":{"name":"ralph","description":"Persistent completion mode. Use when the user explicitly says `/ralph` or clearly wants you to keep iterating until the task is actually finished, repeating implement-verify-fix loops instead of stopping at partial progress."},"skills_sh_url":"https://skills.sh/dgk-dev/dgk-gpt/ralph"},"updatedAt":"2026-05-02T00:57:07.693Z"}}