modify documents

This commit is contained in:
NINI
2026-05-01 02:29:30 +09:00
parent 4b89f4aa96
commit e99b5b8eff
65 changed files with 2814 additions and 72 deletions
+28
View File
@@ -0,0 +1,28 @@
---
description: Draft or update an ADR for a FESA architecture or numerical decision.
---
# /adr
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, `/docs/ARCHITECTURE.md`, `/docs/ADR.md`, and the topic-specific design docs.
- Use `$fesa-adr-update` when available.
- Identify whether this is a new decision, a revision, or a superseding ADR.
## Task
- Capture context, decision, consequences, alternatives considered, and affected docs.
- Update related docs only when needed to keep decisions consistent.
- Record follow-up tasks in `PLAN.md` and completed ADR work in `PROGRESS.md`.
## Verification
- Check links and decision numbering manually.
- Run `python scripts/validate_workspace.py`.
## Result
- **Action**: drafted or updated ADR content
- **Status**: success | partial | blocked
- **Details**: ADR section changed, affected docs, and validation result
@@ -0,0 +1,29 @@
---
description: Guide or validate addition of a new stored reference benchmark.
---
# /benchmark-add
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, `/docs/VERIFICATION_PLAN.md`, `/docs/RESULTS_SCHEMA.md`, and `/docs/ABAQUS_INPUT_SUBSET.md`.
- Use `$fesa-reference-onboarding` when available.
- Do not run Abaqus locally.
## Task
- Help organize the benchmark input, solved reference values such as `*_displacements.csv`, tolerance metadata, unit notes, result fields, and provenance.
- Check that Abaqus keywords are within the Phase 1 subset or explicitly documented as unsupported.
- Ensure comparison values can be tied to step/frame/field/history result paths.
## Verification
- Parse any JSON metadata if present.
- For `*_displacements.csv`, check the required Abaqus displacement/rotation columns.
- Run `python scripts/validate_workspace.py` when files are changed.
## Result
- **Action**: added or validated benchmark artifact
- **Status**: ready | partial | blocked
- **Details**: benchmark name, files, missing metadata, and validation result
@@ -0,0 +1,28 @@
---
description: Review FESA documentation for architecture, numerical, and coordination drift.
---
# /doc-guard
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, `/docs/README.md`, and the docs changed in the current work.
- Use `$fesa-doc-sync` and `$fesa-review` when available.
- If the user explicitly asks for delegated review, use `harness_reviewer` or `progress_plan_auditor`.
## Task
- Check consistency across architecture, ADRs, numerical conventions, Abaqus subset, verification plan, result schema, MITC4 formulation, PLAN.md, and PROGRESS.md.
- Lead with concrete drift findings and file references.
- Keep comments focused on correctness, missing decisions, and validation gaps.
## Verification
- If edits are made, run `python scripts/validate_workspace.py`.
- Confirm that future work stays in PLAN.md and completed work stays in PROGRESS.md.
## Result
- **Action**: reviewed documentation guardrails
- **Status**: success | partial | blocked
- **Details**: findings, edits if any, and validation result
+28
View File
@@ -0,0 +1,28 @@
---
description: Prepare a concise handoff for the next FESA agent session.
---
# /handoff
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, and changed files.
- Use `$fesa-doc-sync` when available.
- Identify what changed, what was verified, what remains blocked, and what should happen next.
## Task
- Add a factual `PROGRESS.md` entry for completed work.
- Update `PLAN.md` only for future tasks, changed ownership, open questions, or blockers.
- Keep the handoff concise enough that a fresh agent can act without private context.
## Verification
- Run `python scripts/validate_workspace.py`.
- Confirm that the next-session required reading remains current.
## Result
- **Action**: prepared handoff
- **Status**: success | partial | blocked
- **Details**: progress entry, plan deltas, blockers, and validation result
@@ -0,0 +1,31 @@
---
description: Draft Harness phase and step files for FESA implementation planning.
---
# /phase-draft
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, `/docs/README.md`, `/docs/HARNESS_ENGINEERING.md`, and all Phase 1 design docs.
- Use `$fesa-readiness` first.
- Use `$fesa-phase-planning` and the repo `harness-workflow` skill when file generation is requested.
- Do not spawn subagents unless the user explicitly asks for delegated work.
## Task
- Identify readiness blockers before drafting steps.
- Keep each phase step self-contained for a fresh Codex session.
- Include a sprint contract: objective, required reading, scope, allowed files, explicit non-goals, tests to write first, reference artifacts, acceptance commands, evaluator checklist, and handoff requirements.
- Do not hide unresolved formulation or reference decisions inside broad implementation steps.
## Verification
- Parse generated JSON phase files.
- Re-read generated `stepN.md` files for self-containment.
- Run `python scripts/validate_workspace.py`.
## Result
- **Action**: drafted phase files
- **Status**: success | partial | blocked
- **Details**: phase directory, step count, blockers, and validation result
@@ -0,0 +1,29 @@
---
description: Reconcile PLAN.md and PROGRESS.md after documentation, planning, or implementation work.
---
# /plan-sync
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, and changed files.
- Use `$fesa-doc-sync` when available.
- Identify completed work, future tasks, blockers, and stale items.
## Task
- Move completed facts into `PROGRESS.md`.
- Keep future tasks and open decisions in `PLAN.md`.
- Preserve history unless the user explicitly asks for archival cleanup.
- Add changed files and verification notes to the progress entry.
## Verification
- Parse or inspect any file format touched.
- Run `python scripts/validate_workspace.py`.
## Result
- **Action**: synchronized plan and progress
- **Status**: success | partial | blocked
- **Details**: entries changed, remaining blockers, validation result
@@ -0,0 +1,28 @@
---
description: Check whether FESA is ready to start Phase 1 implementation planning or coding.
---
# /readiness
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, and `/docs/README.md`.
- Read the readiness-related docs listed by `/docs/README.md`.
- Use `$fesa-readiness` when available.
## Task
- Evaluate each Phase 1 readiness item in `/PLAN.md` and the Implementation Readiness Checklist in `/docs/README.md`.
- Classify each item as ready, blocked, deferred with explicit acceptance, or unknown.
- Do not start implementation from this command.
## Verification
- If files are unchanged, no repository validation is required.
- If readiness notes are written to `PLAN.md` or `PROGRESS.md`, run `python scripts/validate_workspace.py`.
## Result
- **Action**: checked implementation readiness
- **Status**: ready | blocked | partial
- **Details**: readiness table, blockers, and recommended next decision
@@ -0,0 +1,30 @@
---
description: Inspect stored Abaqus reference artifacts for Phase 1 verification readiness.
---
# /reference-check
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, `/docs/VERIFICATION_PLAN.md`, `/docs/RESULTS_SCHEMA.md`, and `/docs/ABAQUS_INPUT_SUBSET.md`.
- Use `$fesa-reference-onboarding` when available.
- If the user explicitly asks for delegated review, use the `reference_artifact_curator` agent.
## Task
- Inspect the `references/` folder when present.
- Check for `.inp` files, `*_displacements.csv` files, tolerance metadata, units notes, Abaqus version/provenance, and mapping to FESA result fields.
- For displacement CSV files, verify the required columns `Node Label`, `U-U1`, `U-U2`, `U-U3`, `UR-UR1`, `UR-UR2`, `UR-UR3`.
- Flag unsupported Abaqus input features as compatibility notes; do not treat stored reference input as automatic Phase 1 parser support.
- Do not run Abaqus locally.
## Verification
- If no files are edited, summarize findings only.
- If manifests or docs are updated, run `python scripts/validate_workspace.py`.
## Result
- **Action**: checked reference artifacts
- **Status**: ready | partial | blocked
- **Details**: artifacts found, missing metadata, and comparison-readiness notes
+28
View File
@@ -0,0 +1,28 @@
---
description: Summarize the current FESA project status and next work from PLAN.md and PROGRESS.md.
---
# /status
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, and `/docs/README.md`.
- Check whether the user asked for status only or also wants edits.
- Do not modify files unless the user explicitly asks for synchronization or cleanup.
## Task
- Summarize current project status, blockers, risks, and the next actionable items.
- Mention whether solver implementation has started.
- Highlight unresolved Phase 1 readiness items and the current validation state.
## Verification
- If no edits were made, state that no validation command was required.
- If the user asked you to update files, run `python scripts/validate_workspace.py`.
## Result
- **Action**: reported FESA status
- **Status**: success | partial | blocked
- **Details**: current objective, blockers, next tasks, and validation note
@@ -0,0 +1,30 @@
---
description: Validate Codex extension files, plugin command metadata, and project documentation metadata.
---
# /verify-docs
## Preflight
- Read `/AGENTS.md`, `/PROGRESS.md`, `/PLAN.md`, and any changed `.codex`, `plugins/`, `.agents/plugins/`, or documentation files.
- Use `$fesa-doc-sync` when available.
## Task
- Parse `.codex/config.toml` and `.codex/agents/*.toml`.
- Parse `.codex/hooks.json`.
- Parse `plugins/*/.codex-plugin/plugin.json`.
- Parse `.agents/plugins/marketplace.json`.
- Check `.codex/skills/*/SKILL.md` frontmatter for `name` and `description`.
- Check `plugins/*/commands/*.md` command Markdown files for description frontmatter.
## Verification
- Run the format checks above.
- Run `python scripts/validate_workspace.py`.
## Result
- **Action**: verified docs and Codex extension files
- **Status**: success | partial | blocked
- **Details**: files checked, failures, and validation result