Install
$ agentstack add skill-xmu-csnoob-praxiskit-seed-to-task-graph ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Seed to Task Graph
Produce a task graph directly from a rough seed. Skips idea and prd artifacts entirely. Most seeds are 1-2 sentences — the gate fires eagerly precisely because there's no intermediate idea/PRD to absorb the drift.
seed -> seed-to-task-graph -> work/task-graph.md + work/SUBAGENT.md -> task-graph-to-batch
Contract
Inputs: seed (raw user input — text, file path, or pasted prompt) Output: work/task-graph.md + work/SUBAGENT.md per schemas/task-graph.schema.md (v3.0); fires references/clarification-gate.md on missing goal/acceptance/forbidden-to-infer fields Preconditions: seed (text or file) provided; seed contains at least an implicit goal (or one is collected via gate) Stop boundary: Does NOT execute tasks, write idea.md or PRD.md, or invoke other transforms.
When to Use This vs seed-to-idea
Light recipe (this skill) — script, prototype, refactor, bug fix; user can directly state acceptance criteria; no external stakeholders need an idea/PRD.
Standard recipe (seed-to-idea → idea-to-prd) — multiple stakeholders, acceptance criteria require derivation from product goals, surface area benefits from explicit non-goals.
The user picks the recipe; this transform does NOT auto-classify.
Workflow
- Load the seed. If user provides a path, read it. If pasted text, use it.
- Load gate reference + schemas. Read
references/clarification-gate.md,schemas/seed.schema.md, andschemas/task-graph.schema.md. - Triage seed length. A 1-2 sentence seed means most goals/acceptance signals are gaps. Apply the Sparse-Input UX Rule from clarification-gate: gate first, draft after.
- Triage seed against
seed.schema.md. Identify which triage dimensions are present. - Apply cross-field rules from
idea.schema.md(distributed → consistencymodel; caching with metadata → cachemetadata_fields). Add triggered fields to gaps. - Required gate fields:
goal_statementif not derivable from seedacceptance_signalonly if the seed does not already provide observable done criteria. If present, copy/quote/split the seed's done criteria into task acceptance criteria and annotate[user].- Any forbidden-to-infer field triggered by cross-field rules
- Call clarification-gate with collected gaps. Prefer host-native multi-question UI (group up to 4).
- Decompose and write artifacts. Follow
references/decompose-task-graph.md(steps 1-10). Frozen contracts step is typically skipped on light recipe.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: xmu-csnoob
- Source: xmu-csnoob/PraxisKit
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.