{"id":"553978d5-108c-440f-bb71-2a5a1a29e1ff","shortId":"8pXGU4","kind":"skill","title":"recall-reasoning","tagline":"Recall the reasoning behind a past change by locating the Claude Code transcript that produced it. Use when the user asks to \"recall reasoning\", \"find reasoning\", \"look up reasoning\", \"recall implementation reasoning\", \"find the rationale\", \"why did I do X\", \"recall from transcri","description":"# Recall Reasoning\n\nLocate the Claude Code transcript that produced a given change and extract the implementer's reasoning. Useful for answering reviewer questions, writing post-hoc explanations, or recovering forgotten context.\n\n## Inputs\n\nAccept any of:\n\n- A commit SHA\n- A file path, optionally with a line number (`<path>:<line>`)\n- A reviewer question plus surrounding context (file and line)\n\nIf only a file is given, `git blame` resolves the commit that last touched the line.\n\n## Step 1: Resolve the Commit and Run the Script\n\nCall `scripts/find_transcript.py` with either `--commit <sha>` or `--file <path>[:<line>]`. Pass `--cwd /path/to/repo` when searching a different repo than the current working directory.\n\n```bash\npython3 <skill-dir>/scripts/find_transcript.py --file <path>:<line>\npython3 <skill-dir>/scripts/find_transcript.py --commit <sha>\n```\n\nThe script:\n\n1. Resolves the commit via `git rev-parse` or `git blame`\n2. Looks up `~/.claude/projects/<encoded-cwd>/` (slashes and dots become dashes)\n3. Ranks candidate transcripts whose mtime is within `--window-days` of the commit (default 14)\n4. Scores candidates by mentions of touched files and tool-use edits on them\n5. Extracts cleaned user prompts and substantive assistant text from the top candidates\n\nThe JSON output has `status`, `commit`, `project_dir`, and `candidates` with `session_id`, `score`, `match_reasons`, and `excerpts`.\n\nStatus values:\n\n- `ok` — candidates returned\n- `no-commit` — couldn't resolve a commit\n- `no-transcripts` — no Claude Code transcript dir for this repo\n- `no-match` — transcripts exist but none match the touched files in the window\n\n## Step 2: Read and Synthesize\n\nIf `status` is `ok`:\n\n1. Start with the top candidate (highest score). Read its excerpts first.\n2. If the excerpts already explain the change, stop. If they are thin or ambiguous, read the full transcript at `jsonl_path` directly for more context.\n3. Ignore candidates with low scores or scores far below the top — they are false positives.\n4. When the reasoning spans multiple excerpts, quote the most specific one.\n\nSynthesize a concise summary tied to the question being answered:\n\n- Lead with the **why**. The diff already shows the what.\n- Quote the implementer's own words when they already say it well.\n- Keep it to one or two paragraphs. Don't narrate the whole session.\n\nIf `status` is anything other than `ok`, report that no reasoning was found and fall back to reading the commit diff and surrounding code. Say so explicitly so it's clear whether the answer still holds up.\n\n## Step 3: Output\n\nReturn the reasoning in this shape:\n\n```\n**Commit:** <short-sha> — <subject>\n**Transcript:** <session-id> (score <N>)\n\n<one or two paragraphs of reasoning, quoting the implementer where useful>\n```\n\nIf no transcript was found:\n\n```\n**Commit:** <short-sha> — <subject>\n**Transcript:** none found (<status>)\n\n<fallback explanation derived from reading the commit and current code>\n```\n\nThen use the TaskList tool and proceed to any remaining task.\n\n## Rules\n\n- Treat excerpts as evidence, not ground truth. The implementer's intent at the time may have changed. If the current code contradicts an excerpt, note the discrepancy.\n- Only read the full `.jsonl` if the excerpts are insufficient. These files are large.\n- Never quote noise prefixes like `<command-message>` or skill-loading stubs. The script filters these out, but stay alert if reading a transcript directly.\n- If multiple candidates have similar high scores, name both and prefer the one whose time window contains the commit.","tags":["recall","reasoning","turbo","tobihagemann","agent-skills","claude-code","claude-skills","developer-tools","skills"],"capabilities":["skill","source-tobihagemann","skill-recall-reasoning","topic-agent-skills","topic-claude-code","topic-claude-skills","topic-developer-tools","topic-skills"],"categories":["turbo"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/tobihagemann/turbo/recall-reasoning","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add tobihagemann/turbo","source_repo":"https://github.com/tobihagemann/turbo","install_from":"skills.sh"}},"qualityScore":"0.590","qualityRationale":"deterministic score 0.59 from registry signals: · indexed on github topic:agent-skills · 280 github stars · SKILL.md body (3,573 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":"2026-04-22T00:54:11.090Z","embedding":null,"createdAt":"2026-04-18T22:03:57.894Z","updatedAt":"2026-04-22T00:54:11.090Z","lastSeenAt":"2026-04-22T00:54:11.090Z","tsv":"'/.claude/projects':172 '/path/to/repo':137 '/scripts/find_transcript.py':150,153 '1':120,157,287 '14':193 '2':169,279,299 '3':178,325,436 '4':194,341 '5':209 'accept':80 'alert':537 'alreadi':303,369,381 'ambigu':313 'answer':67,362,431 'anyth':401 'ask':24 'assist':216 'back':413 'bash':148 'becom':176 'behind':7 'blame':110,168 'call':128 'candid':180,196,221,231,243,292,327,545 'chang':10,58,306,495 'claud':14,51,257 'clean':211 'clear':428 'code':15,52,258,421,499 'commit':84,113,123,132,154,160,191,227,247,252,417,444,463,561 'concis':355 'contain':559 'context':78,99,324 'contradict':500 'couldn':248 'current':145,498 'cwd':136 'dash':177 'day':188 'default':192 'diff':368,418 'differ':141 'dir':229,260 'direct':321,542 'directori':147 'discrep':505 'dot':175 'edit':206 'either':131 'evid':482 'excerpt':239,297,302,347,480,502,513 'exist':268 'explain':304 'explan':74 'explicit':424 'extract':60,210 'fall':412 'fals':339 'far':333 'file':87,100,106,134,151,201,274,517 'filter':532 'find':28,36 'first':298 'forgotten':77 'found':410,462,466 'full':316,509 'git':109,162,167 'given':57,108 'ground':484 'high':548 'highest':293 'hoc':73 'hold':433 'id':234 'ignor':326 'implement':34,62,375,455,487 'input':79 'insuffici':515 'intent':489 'json':223 'jsonl':319,510 'keep':385 'larg':519 'last':115 'lead':363 'like':524 'line':92,102,118 'load':528 'locat':12,49 'look':30,170 'low':329 'match':236,266,271 'may':493 'mention':198 'mtime':183 'multipl':346,544 'name':550 'narrat':394 'never':520 'no-commit':245 'no-match':264 'no-transcript':253 'nois':522 'none':270,465 'note':503 'number':93 'ok':242,286,404 'one':352,388,447,555 'option':89 'output':224,437 'paragraph':391,450 'pars':165 'pass':135 'past':9 'path':88,320 'plus':97 'posit':340 'post':72 'post-hoc':71 'prefer':553 'prefix':523 'proceed':473 'produc':18,55 'project':228 'prompt':213 'python3':149,152 'question':69,96,360 'quot':348,373,453,521 'rank':179 'rational':38 'read':280,295,314,415,507,539 'reason':3,6,27,29,32,35,48,64,237,344,408,440,452 'recal':2,4,26,33,44,47 'recall-reason':1 'recov':76 'remain':476 'repo':142,263 'report':405 'resolv':111,121,158,250 'return':244,438 'rev':164 'rev-pars':163 'review':68,95 'rule':478 'run':125 'say':382,422 'score':195,235,294,330,332,446,549 'script':127,156,531 'scripts/find_transcript.py':129 'search':139 'session':233,397 'sha':85 'shape':443 'show':370 'similar':547 'skill':527 'skill-load':526 'skill-recall-reasoning' 'slash':173 'source-tobihagemann' 'span':345 'specif':351 'start':288 'status':226,240,284,399 'stay':536 'step':119,278,435 'still':432 'stop':307 'stub':529 'substant':215 'summari':356 'surround':98,420 'synthes':282,353 'task':477 'tasklist':470 'text':217 'thin':311 'tie':357 'time':492,557 'tool':204,471 'tool-us':203 'top':220,291,336 'topic-agent-skills' 'topic-claude-code' 'topic-claude-skills' 'topic-developer-tools' 'topic-skills' 'touch':116,200,273 'transcri':46 'transcript':16,53,181,255,259,267,317,445,460,464,541 'treat':479 'truth':485 'two':390,449 'use':20,65,205,457,468 'user':23,212 'valu':241 'via':161 'well':384 'whether':429 'whole':396 'whose':182,556 'window':187,277,558 'window-day':186 'within':185 'word':378 'work':146 'write':70 'x':43","prices":[{"id":"5869f06d-84c5-4ac6-87b8-5bc6d4a3d799","listingId":"553978d5-108c-440f-bb71-2a5a1a29e1ff","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"tobihagemann","category":"turbo","install_from":"skills.sh"},"createdAt":"2026-04-18T22:03:57.894Z"}],"sources":[{"listingId":"553978d5-108c-440f-bb71-2a5a1a29e1ff","source":"github","sourceId":"tobihagemann/turbo/recall-reasoning","sourceUrl":"https://github.com/tobihagemann/turbo/tree/main/skills/recall-reasoning","isPrimary":false,"firstSeenAt":"2026-04-18T22:03:57.894Z","lastSeenAt":"2026-04-22T00:54:11.090Z"}],"details":{"listingId":"553978d5-108c-440f-bb71-2a5a1a29e1ff","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"tobihagemann","slug":"recall-reasoning","github":{"repo":"tobihagemann/turbo","stars":280,"topics":["agent-skills","claude-code","claude-skills","developer-tools","skills"],"license":"mit","html_url":"https://github.com/tobihagemann/turbo","pushed_at":"2026-04-21T12:22:12Z","description":"A composable dev process for Claude Code, packaged as modular skills.","skill_md_sha":"1a47de49209f70c7ea485a6ecb89c752f3f23456","skill_md_path":"skills/recall-reasoning/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/tobihagemann/turbo/tree/main/skills/recall-reasoning"},"layout":"multi","source":"github","category":"turbo","frontmatter":{"name":"recall-reasoning","description":"Recall the reasoning behind a past change by locating the Claude Code transcript that produced it. Use when the user asks to \"recall reasoning\", \"find reasoning\", \"look up reasoning\", \"recall implementation reasoning\", \"find the rationale\", \"why did I do X\", \"recall from transcripts\", or \"find the transcript for this commit\"."},"skills_sh_url":"https://skills.sh/tobihagemann/turbo/recall-reasoning"},"updatedAt":"2026-04-22T00:54:11.090Z"}}