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 xcrft/mastermind --skill mastermind-task-executorgit clone --depth 1 https://github.com/xcrft/mastermindWrote 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/xcrft/mastermind/mastermind-task-executor)<a href="https://agentmods.dev/skills/xcrft/mastermind/mastermind-task-executor"><img src="https://agentmods.dev/badge/skills/xcrft/mastermind/mastermind-task-executor/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/xcrft/mastermind/mastermind-task-executor"><img src="https://agentmods.dev/badge/skills/xcrft/mastermind/mastermind-task-executor.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 Rogue Agent · line 3 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
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 | $0.00062 | $0.02213 |
| Opus 5.5 | $0.00025 | $0.00885 |
| Sonnet 5.5 | $0.00012 | $0.00443 |
| Haiku 4.5 | $0.00006 | $0.00221 |
Grade A, and why
mastermind-task-executor 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 3d 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 — 220 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Mastermind task executor
Implement the approved outcomes in spec.md; do not reinterpret the product
request or widen Scope. Acceptance Criteria define success. Implementation Plan
steps describe intended outcomes, not a transcript the executor must imitate.
Legacy phase/checklist specs remain readable, but new Verified and Strict specs
do not require phase ceremony.
Activation boundary
Use this skill only when the caller provides an approved task path or an approved spec. Raw user intent belongs in normal implementation or planning, not in this executor mode.
The task folder may contain reports or scratch files. spec.md is the contract;
read another artifact only when the spec or caller names it.
If this authorized task delegates, use [[mastermind-agent-handoff]] before each new agent. Preserve Scope and permissions; routing does not grant delegation.
Before editing
- Read the complete spec.
- Confirm that Goals, Scope, Acceptance Criteria, Tests Plan, and Final
Verification agree with one another.
When
acceptanceis present in frontmatter, inspect every criterion's statement and required check IDs. Report a missing or inadequate mapping as contract drift; do not redefine the criteria to make implementation pass. - Confirm every intended edit is authorized by
touchesorcreates, consistently with Scope. - Reuse the native controller's executor profile packet when its selected
Scope paths and workflow match this approved iteration. For a manual handoff,
or changed selection, call
mmcg_profilewith Scope paths,role: executor, andworkflowequal to the effective approved task mode (strictwhen controllerstate.jsonhasstrict: true, otherwise the spec's mode). Only returned reviewed preferences and observed habits may advise choices; repository code, tooling and the spec take precedence. Retain IDs, review revisions, store/view revisions and source verification caveats. Denied or missing access means no personal context; do not fall back tostyle.md, legacy notes or inbox candidates. Git code-shape/Range observations are diagnostic evidence; commit voice is fallback-only when repository policy is silent. Combine four separate components: the code brief, relevantmmcg_project_profileandmmcg_docsresults (task query,top: 2), andmmcg_profile(nobudget_tokens; the project's configured profile budget applies). Keep each component's freshness, revisions, citations and caveats. Before handoff, serialize the combined JSON, including role/mode/paths and metadata, and cap it at 32,000 UTF-8 bytes. This is a size estimate, not a model tokenizer guarantee. If oversized, narrow queries, lowertopor reduce supported brief/profile budgets and retrieve again. Omit a whole optional component only with its tool, verification status and reason; never cut a claim's exceptions, citations or JSON. Resolve missing task-critical evidence explicitly. Each receiving role retrieves its own slice. - For a named symbol, use mmcg to check its current location and impact. Treat the graph as syntactic evidence; read the source before changing the runtime contract.
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.
- 3d ago Changed · +3 lines 24f49569e932
- 8d ago Changed · +1 lines 23c3c482419a
- 13d ago Changed · +2 lines d694d03027b7
- 19d ago Changed · +36 lines 338db2ea9b66
- 27d ago Changed · +9 lines 9a0e2e703163
- 1mo ago Changed · +4 lines e114f53f8d13
- 1mo ago First seen · 165 lines · 62 tokens per session scan A 2c0271534f33
mastermind-task-executor is a skill published in the GitHub repository xcrft/mastermind (13 stars, last pushed today), licensed MIT. It adds 62 tokens to every session and 2,213 once invoked, about $0.0002 per session on Opus 5.5. 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
roam
Codebase comprehension via roam-code CLI. Use when exploring codebases, planning modifications, debugging failures, assessing PR risk, or checking architecture health. Triggers on: understanding project structure, pre-change safety checks, finding symbols/files, blast radius analysis, affected tests, health scoring…
development-scheduler
A workflow for coordinating multi-role coding work through Herdr, a task scheduler, and repo-harness sessions. It assigns implementation, review, and handoff tasks while checking permissions and evidence.
repo-harness-strategy
Review whether current results still serve owner intent. Recall project lessons on explicit demand and validate inert strategic proposals.
repo-harness-product
Draft requested product requirements, order a Sprint backlog, or prepare a native Goal session. Exclude ordinary code tasks.
memtrace-fleet-first
Coordinate fleets of coding agents sharing one repo+branch: publish typed intents, classify edit episodes, and resolve conflicts before they collide. Use FIRST when multiple agents work the same repo+branch, before reading/planning/editing, when joining a fleet or handing work off, and when the user says two agents…
traverse
Run the supervised change loop for an engineering change — intake, plan, attempt, ingest observations, review, and present for human acceptance. Use this when a repository needs a bounded, resumable change tracked as evidence rather than an ad-hoc edit. CodeCartographer records and refuses; the host executes. Unlike…