Jeden Skill in Manus ausführen
mit einem Klick

Jeden Skill in Manus mit einem Klick ausführen

orcho-prompt-engine

Sterne3

Forks0

Aktualisiert30. Juni 2026 um 16:45

Use for Orcho prompt authoring/rendering internals: orcho-core/pipeline/prompts/*, orcho-core/core/_prompts/*, prompt templates/contracts, cache-aware rendering, session/delta prompt metadata, prompt snapshots, and prompt boundary tests. Pair with evidence/observability only when rendered prompt data is stored as evidence/events/artifacts.

Installation

Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.

In Manus ausführen

Quelle

symphos-ai

symphos-ai/orcho-core

GitHub-Repository öffnen Creator-Repositorys ansehen

Download

In Manus ausführen

Verwandte BerufeSOC

Basierend auf der SOC-Berufsklassifikation

SoftwareentwicklerInformatik- und Mathematikberufe·SOC 15-1252

SKILL.md

readonly

name	orcho-prompt-engine
description	Use for Orcho prompt authoring/rendering internals: orcho-core/pipeline/prompts/, orcho-core/core/_prompts/, prompt templates/contracts, cache-aware rendering, session/delta prompt metadata, prompt snapshots, and prompt boundary tests. Pair with evidence/observability only when rendered prompt data is stored as evidence/events/artifacts.

Orcho Prompt Engine

Work safely on prompt composition, prompt contracts, cache layout, and prompt evidence. Keep machine contracts in code-owned prompt blocks, not loose prose.

First Reads

orcho-core/AGENTS.md
orcho-core/pipeline/prompts/AGENTS.md
orcho-core/core/_prompts/AGENTS.md
orcho-core/docs/adr/0009-composable-prompt-parts.md
orcho-core/docs/adr/0028-cache-first-physical-wire-layout.md
changed prompt builder/template files

Owns

pipeline/prompts/*
core/_prompts/*
prompt part ordering and cache metadata
prompt contract templates and protected system-tail policy
prompt render evidence and prompt snapshots

Does Not Own

runtime adapter invocation -> orcho-core-runtime-session
SDK/MCP public payload shape -> orcho-core-sdk-wire
parser/gate verdict semantics -> orcho-core-quality-gates

Invariants

Think in ordered prompt parts, not one rendered string.
Preserve cache-stable prefix ordering unless deliberately changing it with tests.
Put dynamic facts in typed prompt parts with explicit stability/cache scope.
Keep protected contracts in code-owned templates.
Do not silently widen editable markdown into protocol policy.

Verification

From orcho-core: python -m pytest -q tests/unit/pipeline/prompts
From orcho-core: python -m pytest -q tests/unit/pipeline/runtime/test_snapshot_session_parity.py when cache/session behavior changes
Review prompt snapshot/golden diffs before accepting them.

Neighbor Skills

orcho-core-runtime-session when prompt/session continuity changes
orcho-core-evidence-observability when rendered prompt data is persisted as evidence
orcho-lab when evaluating prompt variants

Mehr aus diesem Repository

gleiches Repository

orcho-core-cli-ux

symphos-ai/orcho-core

Use when editing Orcho CLI behavior: orcho-core/cli/*, CLI help text, output modes, interactive prompts, stdout/stderr formatting, transcript behavior, latest/resume commands, or public `orcho`, `orcho run`, `orcho cross`, `orcho status`, `orcho evidence`, `orcho diff`, `orcho web` UX. Do not use for pipeline semantics alone.

2026-06-303

orcho-core-cross-project

symphos-ai/orcho-core

Use when editing cross-project orchestration lifecycle and wiring: orcho-core/pipeline/cross_project/*, handoff, contract check, artifact bundle, profile projection, child dispatch, cross-project gate handoff, release-gate orchestration, and final-acceptance bundle flow. Pair with quality-gates for parser/verdict/gate semantics; do not use for single-project gate/parser changes.

2026-06-303

orcho-core-evidence-observability

symphos-ai/orcho-core

Use for Orcho evidence/observability storage, emission, mirrors, bundles, events, metrics, run diff, and artifact metadata: orcho-core/pipeline/evidence/*, pipeline/observability/*, core/observability/*, pipeline/artifacts/types.py, run_diff.py, artifact_mirror.py, run_logging.py, including persisted prompt-render evidence. Do not use for prompt wording/templates/render contracts unless paired.

2026-06-303

orcho-core-isolation-worktrees-sandbox

symphos-ai/orcho-core

Use when editing orcho-core isolation and workspace safety: pipeline/project/isolation_setup.py, worktree selection/bootstrap, pre-run dirty intake, sandbox policy, runspace paths, dependency checkout subjects, git worktree tests, sandbox ADRs, or generated worktree handling. Do not use for runtime agent sessions or phase lifecycle unless paired.

2026-06-303

orcho-core-phases-engine

symphos-ai/orcho-core

Use when editing phase lifecycle/execution mechanics: phase handlers/registry, orcho-core/pipeline/phases/*, project run phase bookkeeping, subtask DAG execution, StepOutcome, subtask receipts, done-criteria attestation, phase transitions, and phase-level handoff/routing. Do not use for parser verdicts, gate decisions, or runtime adapter internals unless paired.

2026-06-303

orcho-core-profiles-plugins

symphos-ai/orcho-core

Use when editing Orcho profiles and plugin extension surfaces: profile schema/catalogue, semantic profiles, operating modes, profile projection, phase config, profile-driven gate scheduling/configuration, PluginConfig loading, entry-point groups for agent_runtimes/phases/skills, plugin authoring docs, and profile validation. Do not use for runtime execution internals or SDK wire unless paired.

2026-06-303

name	orcho-prompt-engine
description	Use for Orcho prompt authoring/rendering internals: orcho-core/pipeline/prompts/, orcho-core/core/_prompts/, prompt templates/contracts, cache-aware rendering, session/delta prompt metadata, prompt snapshots, and prompt boundary tests. Pair with evidence/observability only when rendered prompt data is stored as evidence/events/artifacts.

Orcho Prompt Engine

Work safely on prompt composition, prompt contracts, cache layout, and prompt evidence. Keep machine contracts in code-owned prompt blocks, not loose prose.

First Reads

orcho-core/AGENTS.md
orcho-core/pipeline/prompts/AGENTS.md
orcho-core/core/_prompts/AGENTS.md
orcho-core/docs/adr/0009-composable-prompt-parts.md
orcho-core/docs/adr/0028-cache-first-physical-wire-layout.md
changed prompt builder/template files

Owns

pipeline/prompts/*
core/_prompts/*
prompt part ordering and cache metadata
prompt contract templates and protected system-tail policy
prompt render evidence and prompt snapshots

Does Not Own

runtime adapter invocation -> orcho-core-runtime-session
SDK/MCP public payload shape -> orcho-core-sdk-wire
parser/gate verdict semantics -> orcho-core-quality-gates

Invariants

Think in ordered prompt parts, not one rendered string.
Preserve cache-stable prefix ordering unless deliberately changing it with tests.
Put dynamic facts in typed prompt parts with explicit stability/cache scope.
Keep protected contracts in code-owned templates.
Do not silently widen editable markdown into protocol policy.

Verification

From orcho-core: python -m pytest -q tests/unit/pipeline/prompts
From orcho-core: python -m pytest -q tests/unit/pipeline/runtime/test_snapshot_session_parity.py when cache/session behavior changes
Review prompt snapshot/golden diffs before accepting them.

Neighbor Skills

orcho-core-runtime-session when prompt/session continuity changes
orcho-core-evidence-observability when rendered prompt data is persisted as evidence
orcho-lab when evaluating prompt variants