ralph-specum-start
This skill should be used only when the user explicitly asks to use `$ralph-specum-start`, or explicitly asks Ralph Specum in Codex to start or resume a spec.
Install / Use
npx skills add tzachbon/smart-ralph --skill ralph-specum-startInstalls into whichever agent you are using.
SKILL.md
Installable skill definition
Quality Score
Category
LegalSupported Platforms
Tags
Our assessment of ralph-specum-start
ralph-specum-start scores 81/100 on our quality scale, 183rd of 210 Legal skills we index.
Its SKILL.md is 8.5 KB long, split into 6 sections and no code examples: a thorough specification that gives an agent plenty to work with.
It has 550 GitHub stars, a meaningful sign that others use it.
Maintenance, license and trust
- The repository was last updated about 2 months ago, so ralph-specum-start is actively maintained.
- Our last check on 2026-11-08 found the source still online.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 100/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-10-05. Automated pattern scan on 2026-10-04. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
ralph-specum-start compared with similar skills
All 4 of these similar skills score higher than ralph-specum-start; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| ralph-specum-start (this skill)by tzachbon | 81 | 550 | 2mo ago | SKILL.md |
| algorithmic-artby anthropics | 100 | 177.9k | 57d ago | SKILL.md |
| pptxby anthropics | 100 | 177.9k | 57d ago | SKILL.md |
| prospectingby coreyhaines31 | 100 | 60.1k | 7d ago | SKILL.md |
| ad-creativeby coreyhaines31 | 100 | 60.1k | 7d ago | SKILL.md |
Frequently asked questions
- How do I install ralph-specum-start?
- Run
npx skills add tzachbon/smart-ralph --skill ralph-specum-start. The install tabs above show the steps for each supported agent. - Which AI agents does ralph-specum-start work with?
- It is written for Claude Code and OpenAI Codex, as a SKILL.md file. Other agents that read the same format can often use it too.
- Is ralph-specum-start safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 100/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is ralph-specum-start still maintained?
- The repository was last updated about 2 months ago, so ralph-specum-start is actively maintained.
Skill content
View source on GitHubname: ralph-specum-start
description: This skill should be used only when the user explicitly asks to use $ralph-specum-start, or explicitly asks Ralph Specum in Codex to start or resume a spec.
metadata:
surface: helper
action: start
Ralph Specum Start
Use this for the start and new entrypoints.
Derive RALPH_CODEX_PLUGIN_ROOT from this loaded skill by resolving two parent directories from the SKILL.md directory. Never derive it from the project working directory.
Contract
- Read
.claude/ralph-specum.local.mdwhen present - Default specs root is
./specs - Keep
.current-specin the default specs root - Keep the standard Ralph files stable
- Merge
.ralph-state.json. Do not replace the full object
Action
- Parse explicit name, goal, exact
--quickor exact--interactive,--resume <prototype-id>, commit flags, optional specs root, and optional--tasks-size fine|coarse. Reject both mode flags together,-q, variants, and natural-language substitutes. - Classify new versus resume intent, then resolve the classified target by explicit path, exact name, or
.current-spec. A new-spec request does not inherit or recover the existing current spec. - If the same name exists in multiple configured roots, stop and require a full path.
- Check active epic context from
specs/.current-epicwhen no explicit spec was chosen. - For large or cross-cutting goals, route to triage instead of forcing a single spec.
newis an alias here. Create the spec directory if needed.- Initialize or merge state with:
source: "spec"name- exact
goaltext used by the context gate basePathphase: "research"taskIndex: 0totalTasks: 0taskIteration: 1maxTaskIterations: settings default or 5globalIteration: 1maxGlobalIterations: 100commitSpec: settings auto_commit_spec or truerelatedSpecs: []awaitingApproval: falsewhile the goal interview is active- preserve or set
quickMode - preserve or set
granularitywhen--tasks-sizewas supplied - preserve or set
epicNamewhen starting from an epic suggestion
- Update
.current-spec. - Write
.progress.mdwith goal, current phase, next step, blockers, learnings, and skill discovery results. - On resume, prefer
tasks.mdand present files over stale state when they disagree. - Run
phase_gate.py modethrough"$RALPH_CODEX_PLUGIN_ROOT/scripts/phase_gate.py"withSTATEand the exact supplied mode flag. No flag normalizes invalid legacy quick state to interactive. - Run skill discovery pass 1 against the goal. Collect plugin skills, project
.agents/skills, project.claude/skills, and the current Codex harness catalog. Always select explicitly named skills and record shadowed duplicates. - Load
"$RALPH_CODEX_PLUGIN_ROOT/skills/interview-framework-codex/SKILL.md", its required algorithm and domain-modeling references, and every selected domain contract in both interactive and quick mode. In interactive mode, follow that algorithm for the start goal territory: outcome and observable success, material scope boundaries, critical constraints, and viable high-level approach. Ask no spec-path, branch, task-size, or discoverable question. - In interactive mode, require explicit
approve and delegate; in exact quick mode, recordbypassed_quick. In both modes, runphase_gate.py check-delegationwith the current loaded-manifest identity before creating aresearch-analystchild. Apply the shared hard-transition invariant before fresh or resumed research dispatch. A failedcheck-delegationin either mode stops this invocation before phase transition, child dispatch, or target-artifact write; after a normal-mode failure, only the next explicit invocation creates a fresh manifest/interview identity. After an exact--quickdelegation failure, the next explicit invocation reruns discovery, records a freshphaseSkillLoadand interview identity, and does not reuse a terminalbypassed_quickinterview or its discovery revision. Only a matching in-progresscollectingorawaiting_confirmationinterview is resumable. Exact--quickretains its existing question-and-approval bypass and discovery, manifest, delegation, and writer checks. - Keep the existing child packet, identity tuple, and receipt behavior unchanged: pass the absolute helper path, state path, identity tuple, a unique teammate dispatch identity, and the verbatim
phaseSkillLoadmanifest. The child must record its loads and passcheck-agent-writewith that unique identity before writingresearch.md. - Validate
research.md, mergephase: "research"andawaitingApproval: truein interactive mode orfalsein exact quick mode, update progress, and present artifact approval when interactive. - In exact quick mode, record the quick bypass, generate missing artifacts in order, skip normal approval pauses, and continue into implementation in the same run.
Prototype Reconciliation and Resume
After resolving the classified target and before normal resume or quick routing. Skip existing-current-spec recovery for new-spec intent:
- Use
"$RALPH_CODEX_PLUGIN_ROOT/scripts/resolve_spec_paths.py"and only its resolvedbasePath. When bothbasePathand<basePath>/.ralph-state.jsonexist, run"$RALPH_CODEX_PLUGIN_ROOT/scripts/prototype_records.py" reconcile --base-path "$BASE_PATH" --state "$BASE_PATH/.ralph-state.json", then re-read state. - Treat a missing
activePrototypesfield as an empty map. Sort entries bycreated, then ID. - Resume an explicit active ID through
$ralph-specum-prototype --resume <id>and stop this skill. In normal mode, resume the sole active entry automatically. When several remain, list deterministic IDs with question, status, blocker,returnPhase, andreturnTaskIndex, then stop for an explicit ID. - In quick mode, ask no question. Sort entries that block design by
created, then ID. At the post-requirements boundary, route through$ralph-specum-prototype --quick; the prototype skill takes over the oldest design blocker and owns every decision. Preserve earlier quick flow and unrelated entries. - Treat each
resume_reviewcandidate as recovery work, not terminal evidence. Parse the exact candidate, verify its ID and hash, and reconstruct a minimal recovery entry from the record and source pointers. Before reviewer dispatch, reserve its ID through create-onlylocked_state.py upsert-prototypewithstatus: reviewing, the exactcandidateHash, null owner and lease fields, a null or absentharnessRun.id, blocker and return fields, source pointers, and recovery timestamps. This is a recovery-only entry: cancellation verifies thatowner,leaseToken, andharnessRun.idare all null or absent and that no builder is associated, then skips interrupt and release; inconsistent builder ownership fails closed. On a concurrent reservation, re-read and continue only if itscandidateHashmatches; never overwrite it. Route the restored entry through deterministic exact-candidate review and publish recovery. If reconstruction or reservation cannot proceed, stop and report the candidate ID and candidate hash. Exclude quarantined or malformed records. If no overlay exists, continue the existing start behavior without extra output. - Quick mode completes its automatic local route and does not reach the interactive Response Handoff below.
Branch Isolation
- If the user wants isolation, offer a feature branch in place or a worktree with a feature branch.
- If a worktree is created, stop after creation and ask the user to continue from that worktree.
Response Handoff
- After creating or resuming the spec, name the resolved spec path and summarize the current state briefly.
- After setup in normal mode, begin the goal grill in the same run without stopping at a setup prompt.
- After final
approve and delegate, dispatch research in the same turn. - After
research.mdin interactive mode, name the file, summarize it, and end with exactly one artifact approval prompt:approve current artifactrequest changescontinue to requirements
- With exact
--quick, do not show this prompt; continue directly through the remaining phases after the gates succeed. - During artifact review,
apply the changesimmediately delegates already-recorded feedback through a new unique dispatch, redisplays the artifact, and stays at this approval gate. Ask one focused change question only when no feedback is pending. Control-onlycontinue,proceed, andgo aheadapprove nothing.
Related Skills
algorithmic-art
177.9kCreating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems.
pptx
177.9kUse this skill any time a .pptx or .potx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx or .potx file (even if the extracted content will be used elsewhere, like in an em…
prospecting
60.1kWhen the user wants to find and qualify prospects for outbound, or asks who to target first, which signal to prioritize, or where to spot buyers with the problem: buying signals, enrichment, verification, account research.
ad-creative
60.1kWhen the user wants to write or judge ad creative, or asks which ad, hook, or angle to run or kill: headlines, primary text, and variations for Google, Meta, LinkedIn, and TikTok, fresh angles when creative fatigues. Test hypotheses name the metric, comparison, and expected direction.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
