Name: telperion
Author: DanielKillenberger

Imported from DanielKillenberger/telperion (AGENTS.md). Install upstream with npx skills add DanielKillenberger/telperion. Copyright stays with the author.

Flow-Next

This project uses Flow-Next for ALL task tracking. flowctl comes from the flow-next plugin install — every flow-next skill resolves it itself, and on Claude Code it is also on PATH. Do NOT create markdown TODOs or use TodoWrite. Cold session: flowctl brief first — one bounded call (specs, ready tasks, memory); go deeper with show/cat/anchor <task-id>.

Lifecycle: flowctl list / show fn-N.M / start fn-N.M / done fn-N.M --summary-file s.md --evidence-json e.json (e.json: {"commits": ["<sha>"], "tests": ["<cmd>"], "prs": []})
BEFORE any other flowctl operation, or when unsure of a flag: run flowctl usage (CLI cheatsheet + orchestration recipes) or flowctl --help.
BEFORE bridging work to another model/CLI (codex exec, cursor-agent, claude -p, grok) or picking an implementation/review model: run flowctl usage and follow "Orchestration & model steering" exactly.
Creating a spec: write it directly — /flow-next:plan is task breakdown only. flowctl spec create --title "Short title" --plan-file plan.md --json, then /flow-next:plan <spec-id>. Scaffold cascade (first match wins): SPEC.md -> spec.md -> bundled template.
Substantial replies (reports, reviews, multi-section answers): invoke /flow-next:prose BEFORE drafting — the artifact prose contract applies to chat replies too. Short conversational turns skip it.
If flowctl is not found: your shell lacks the plugin's scripts/ dir on PATH (only Claude Code injects it). Resolve it the way the skills do - the plugin install's scripts/flowctl (Claude/Droid: plugin-root env var; Codex: ${CODEX_HOME:-$HOME/.codex}/scripts/flowctl; Cursor/Grok: two levels above any flow-next SKILL.md) - or update/reinstall the flow-next plugin. A repo with no .flow/ yet: run /flow-next:setup.

Model routing

{ "ocm": "1", "id": "danielkillenberger-telperion-telperion", "kind": "skill", "name": "telperion", "description": "This project uses Flow-Next for ALL task tracking. `flowctl` comes from the flow-next plugin install — every flow-next skill resolves it itself, and on Claude Code it is also on PATH. Do NOT create markdown TODOs or use TodoWrite. Cold session: `flowctl brief` first — one bounded call (specs, ready tasks, memory); go deeper with `show`/`cat`/`anchor <task-id>`.", "publisher": "DanielKillenberger", "version": "1.0.0", "capabilities": { "domains": [ "coding" ], "tags": [ "skill-md", "github-agents" ], "languages": [ "en" ] }, "quality_prior": 0.6, "examples": [ "This project uses Flow-Next for ALL task tracking. `flowctl` comes from the flow-next plugin install — every flow-next skill resolves it itself, and on Claude Code it is also on PATH. Do NOT create markdown TODOs or use TodoWrite. Cold session: `flowctl brief` first — one bounded call (specs, ready tasks, memory); go deeper with `show`/`cat`/`anchor <task-id>`." ], "primary": false, "metadata": { "source": { "provider": "github-agents", "repository": "https://github.com/DanielKillenberger/telperion", "path": "AGENTS.md", "ref": "2f33981a3107f39e1bb2b0603de7ac4e4a417558", "url": "https://github.com/DanielKillenberger/telperion/blob/2f33981a3107f39e1bb2b0603de7ac4e4a417558/AGENTS.md", "key": "DanielKillenberger/telperion/AGENTS.md" } }, "instructions": "\n\n## Flow-Next\n\nThis project uses Flow-Next for ALL task tracking. `flowctl` comes from the flow-next plugin install — every flow-next skill resolves it itself, and on Claude Code it is also on PATH. Do NOT create markdown TODOs or use TodoWrite. Cold session: `flowctl brief` first — one bounded call (specs, ready tasks, memory); go deeper with `show`/`cat`/`anchor <task-id>`.\n\n- Lifecycle: `flowctl list` / `show fn-N.M` / `start fn-N.M` / `done fn-N.M --summary-file s.md --evidence-json e.json` (e.json: `{\"commits\": [\"<sha>\"], \"tests\": [\"<c", "cost": { "context_tokens": 921 } }

telperion

About

Flow-Next

Model routing

Use it

Manifest

Reviews