Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add tzachbon/smart-ralph --skill ralph-specum-tasksgit clone --depth 1 https://github.com/tzachbon/smart-ralphWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/tzachbon/smart-ralph/ralph-specum-tasks)<a href="https://agentmods.dev/skills/tzachbon/smart-ralph/ralph-specum-tasks"><img src="https://agentmods.dev/badge/skills/tzachbon/smart-ralph/ralph-specum-tasks/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/tzachbon/smart-ralph/ralph-specum-tasks"><img src="https://agentmods.dev/badge/skills/tzachbon/smart-ralph/ralph-specum-tasks.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Agent Snooping · line 38 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00043 | $0.01388 |
| Opus 5 | $0.00022 | $0.00694 |
| Sonnet 5 | $0.00009 | $0.00278 |
| Haiku 4.5 | $0.00004 | $0.00139 |
Grade A, and why
ralph-specum-tasks scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ralph Specum Tasks
You are a coordinator, not a task planner -- delegate ALL work to a task-planner sub-agent.
Derive RALPH_CODEX_PLUGIN_ROOT from this loaded skill by resolving two parent directories from the SKILL.md directory. Never derive it from the project working directory.
Contract
- Resolve the active spec by explicit path, exact name, or
.current-spec - Require
requirements.mdanddesign.md - Merge state fields only
- Keep the Ralph disk contract unchanged
Action
- Resolve the active spec. If none exists, stop.
- Require
requirements.mdanddesign.md. Readresearch.mdwhen present,.progress.md, and current state. - Run
phase_gate.py modethrough"$RALPH_CODEX_PLUGIN_ROOT/scripts/phase_gate.py"withSTATEand exact--quick, exact--interactive, or no flag. Reject both,-q, variants, and natural-language substitutes. - In interactive mode, require artifact approval for the current
design.mdbefore starting tasks. Exact quick mode continues with the validated artifact. - Run prototype record selection with the resolved
basePathbefore generation:python3 "$RALPH_CODEX_PLUGIN_ROOT/scripts/prototype_records.py" select-downstream --base-path "$BASE_PATH" --state "$BASE_PATH/.ralph-state.json" - Include only affected, valid,
gateApproved: true, non-superseded records returned by the selector. Exclude malformed, superseded, skipped, failed, inconclusive, cancelled, and normal-mode excluded records. - Stop before generation when selection reports an
activePrototypesblocker for tasks. Name the active ID and route resume through$ralph-specum-prototype. Allow proven unrelated prototypes when the selector reports no task dependency. - Stop when selection reports stale
design.md, stale task indexes, or a stale upstream artifact that design depends on. Route to the earliest stale phase and do not plan from stale design. - Clear any prior approval gate by merging
awaitingApproval: falsebefore generation. - Respect
granularityfrom state. Allow--tasks-size fine|coarseto override it. Treat task sizing as administration, not an interview question. In exact quick mode, default unset granularity tofine. - When
research.mdexists, require skill discovery pass 2 against the goal plus final research. When it is absent, require pass 1 against the goal alone. Run the applicable pass when the state lacks its revision. Select explicitly named skills and record harness-shadowed duplicates. - Load
"$RALPH_CODEX_PLUGIN_ROOT/skills/interview-framework-codex/SKILL.md", its required algorithm and domain-modeling references, and all selected domain contracts in both interactive and quick mode. In interactive mode, follow the algorithm for critical delivery slicing, dependency order, rollout risk, and verification thresholds. Inspect commands, file layout, and existing test tools instead of asking. - 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 the child. Apply the shared hard-transition invariant before task-planner dispatch. A failedcheck-delegationin either mode ends this invocation before phase transition, child dispatch, or target-artifact write; after a normal-mode failure, the next explicit invocation creates a fresh manifest/interview identity and does not create an alternate quick bypass. 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. - Delegate task planning to a
task-plannersub-agent. Pass the absolute helper path, state path, identity tuple, unique teammate dispatch identity, verbatim manifest, requirements, design, research, selected prototype evidence, the clean blocker/stale-gate result, and interview context. The child reloads and records the manifest, passescheck-agent-writewith that unique identity, and writestasks.md. Do NOT write tasks.md yourself. - Read the sub-agent's output and validate it exists.
- Count tasks and merge state with:
phase: "tasks"awaitingApproval: true(orfalsewhen--quickis active)taskIndex: first incomplete or totalTaskstotalTasks: counted tasks
- Update
.progress.mdwith the phase breakdown, next milestone, blockers, next step, chosen granularity, skill discovery, and verification strategy. - If spec commits are enabled, commit only the spec artifacts.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago Changed 49cfc659d046
- 11d ago First seen · 69 lines · 43 tokens per session scan A 2bf9a354fe90
ralph-specum-tasks is a skill published in the GitHub repository tzachbon/smart-ralph (537 stars, last pushed 7d ago), licensed MIT. It adds 43 tokens to every session and 1,388 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
remember
Review reusable project knowledge and decide what belongs in project memory, notepad, or durable docs.
docx-comment-reply
Reply to comments (批注) in Word .docx/.doc files: extract comment context, draft replies, write threaded replies back, and validate OOXML.
explaining-machine-learning-models
Explain trained machine learning models through feature attribution, local explanations, and behavior summaries. Use as an explicit/manual helper once a model already exists, not for training ownership, leakage auditing, or general ML strategy selection.
config-validator
Validate AIWG configuration files and project setup for correctness and completeness.
aiwg-regenerate-copilot
Regenerate copilot-instructions.md for GitHub Copilot with vendor-specific content only.
steward-prep-delivery
Steward-assisted prep for filing issues and PRs — environment capture, template selection, duplicate detection, delivery-policy compliance check.