Sigil: Spellcraft

• invoke owns early discovery, definition, design, planning, and handoff packets.
• spellcraft owns spell composition, phase/gate design, local installation/adaptation, validation, observability, reflection, and revision.
• sigil-development owns changes to individual sigil contracts referenced by a spell.
• task-session owns bounded execution from approved work-pack tasks or SWUs.
• experiment-harness owns repeatable test mechanics, realistic prompts, live Codex examples, validation reports, and telemetry emission for reusable spells.

When the input is an Invoke handoff, consume the handoff as source context, then take lifecycle ownership of the spell. Do not send the user back to Invoke unless the handoff is missing the workflow objective, target repository scope, or intended output artifact.

When implementation or adaptation work is delegated through Task Session and the codex-goal adapter:

1. Spellcraft owns the lifecycle work-pack and promotion decision.
2. Task Session selects one ready task/SWU and checks gates.
3. The Codex Goal adapter generates or hands off one native /goal.
4. Codex executes the bounded runtime goal.
5. Task Session reviews the result against the original task/SWU and syncs work-pack evidence.
6. Experiment Harness runs or validates the artifact-local examples/regimes.
7. Spellcraft consumes the experiment report, observability signal, and reflection trigger state before marking lifecycle progress.

Required runtime evidence shape:

runtime: codex
adapter: codex-goal
source_swu: <id>
result: pass | flag | block | interrupted
files_touched:
  - <path>
validation:
  - <command or review evidence>
experiment_harness:
  status: pass | flag | block | not_run
  report: <path or none>
remaining_blockers:
  - <blocker or none>
lifecycle_owner_next_step: validate | observe | reflect | iterate | promote

A spell implementation or adaptation SWU is not lifecycle-complete until runtime evidence is reviewed and the relevant experiment harness state is updated or explicitly blocked.

.arcanum/spells/<spell-name>/README.md

For reusable library spells, write to:

spells/<spell-name>/README.md

• canonical ID,
• aliases, when any,
• purpose,
• trigger conditions,
• required sigils,
• optional sigils,
• prerequisites,
• shared state,
• execution phases,
• handoff artifacts,
• gates,
• failure policy,
• local customization,
• observability,
• output contract.

• spell name,
• mode,
• phases attempted,
• sigils invoked,
• gates passed,
• gates blocked,
• handoff failures,
• outputs produced,
• validation result,
• user corrections,
• recommended spell updates.

• keep spells distinct from sigil tiers,
• resolve aliases to stable canonical IDs,
• reference sigils rather than copying their process bodies,
• define phase inputs, outputs, gates, and failure policy,
• preserve local customization without forking upstream sigils,
• validate referenced sigils and handoff artifacts,
• define or preserve an experiment harness for reusable spells,
• treat Codex Goal evidence as SWU execution evidence, not as reusable-behavior validation,
• require experiment harness evidence or a named block before promotion readiness,
• define observability for reusable spells,
• return a clear next action.

• treating spells as a fourth sigil tier,
• allowing one alias to resolve to multiple active spells,
• hiding multiple unrelated workflows in one spell,
• copying full sigil instructions into a spell file,
• installing a library spell without adapting local paths,
• running later phases after an earlier gate blocks,
• letting a spell mutate a sigil contract without explicit maintenance work,
• omitting failure policy for handoff gaps.

## Spellcraft Result

- Mode: design | install | run-plan | validate | observe | reflect
- Spell: <name>
- Canonical ID: <spell-id>
- Alias used: <alias or none>
- Scope: library | local
- Spell file: <path>
- Sigils referenced: <list>
- Phases: <count>
- Validation: pass | block | flag | not run
- Observability: configured | unavailable | not needed
- Next action: <action>