| name | local-journal |
| description | Use this skill to maintain the project-state files under `.agents/` in a Main Sequence repository. This skill owns reconciliation of current state, stable references, next actions, blockers, and historical notes after material work. |
Local Journal And Project State Maintenance
Overview
Use this skill after material work to reconcile the project-state files under .agents/.
This skill owns:
.agents/brief.md.agents/tasks.md.agents/record.md.agents/status.md.agents/journal.md
This Skill Can Do
- update
.agents/brief.md when user intent, scope, or success criteria changed
- update
.agents/tasks.md when next actions, follow-up work, or open blockers changed
- update
.agents/record.md when stable commands, identifiers, paths, or operational notes changed
- update
.agents/status.md when verified state changed
- append to
.agents/journal.md when a milestone, failure, repeated issue, or investigation outcome should be preserved
- keep the five
.agents/ files consistent with each other
This Skill Must Not Claim
This skill must not claim ownership of:
- DataNode semantics
- SimpleTable semantics
- API design
- Command Center payload design
- job and release configuration semantics
- RBAC semantics
- market, VFB, dashboard, or pricing semantics
This skill maintains project-state files. It does not own domain implementation rules.
Route Adjacent Work
- bootstrap and routing:
.agents/skills/project_builder/SKILL.md
- project audits and blocker classification:
.agents/skills/maintenance/bug_auditor/SKILL.md
- DataNodes:
.agents/skills/data_publishing/data_nodes/SKILL.md
- SimpleTables:
.agents/skills/data_publishing/simple_tables/SKILL.md
- FastAPI and application-facing APIs:
.agents/skills/application_surfaces/api_surfaces/SKILL.md
- Command Center workspaces:
.agents/skills/command_center/workspace_builder/SKILL.md
- AppComponents and custom forms:
.agents/skills/command_center/app_components/SKILL.md
- jobs, schedules, resources, releases, and artifacts:
.agents/skills/platform_operations/orchestration_and_releases/SKILL.md
- RBAC and sharing:
.agents/skills/platform_operations/access_control_and_sharing/SKILL.md
- assets and translation tables:
.agents/skills/markets_platform/assets_and_translation/SKILL.md
- Virtual Fund Builder:
.agents/skills/markets_platform/virtualfundbuilder/SKILL.md
- instruments and pricing:
.agents/skills/markets_platform/instruments_and_pricing/SKILL.md
- Streamlit dashboards:
.agents/skills/dashboards/streamlit/SKILL.md
Read First
.agents/AGENTS.md.agents/skills/project_builder/SKILL.md- the domain skill that produced the material change
.agents/brief.md.agents/tasks.md.agents/record.md.agents/status.md.agents/journal.md
Inputs This Skill Needs
Before editing the .agents/ files, collect or infer:
- what materially changed
- what was verified versus assumed
- whether user intent or scope changed
- whether new follow-up tasks were created
- whether stable commands, identifiers, or paths changed
- whether the turn produced a milestone, failure, repeated issue, or investigation result worth preserving
Required Decisions
For every maintenance pass, decide:
- Does
.agents/brief.md need to change?
- Does
.agents/tasks.md need to change?
- Does
.agents/record.md need to change?
- Does
.agents/status.md need to change?
- Does
.agents/journal.md need a new entry?
- Are the files still consistent with each other after the update?
Build Rules
1. Update only after material work
Do not rewrite the .agents/ files for trivial lookups.
Use this skill after:
- implementation work
- verification work
- blocker discovery
- scope change
- documentation mismatch discovery
- new follow-up work discovery
2. Keep verified state separate from assumptions
.agents/status.md must distinguish verified facts from assumptions or pending checks.
3. Keep stable references in .agents/record.md
Put stable commands, ids, file paths, and operational references in .agents/record.md, not in .agents/journal.md.
4. Keep .agents/journal.md append-only in spirit
Use .agents/journal.md for historical notes:
- milestones
- failures
- repeated issues
- investigations
- suspected SDK or platform issues
Do not use it as the current source of truth for next actions or verified state.
5. Keep .agents/tasks.md actionable
Write follow-up work as concrete tasks, not vague notes.
6. Keep .agents/brief.md short and stable
It should reflect the current user goal and success condition, not the full execution history.
Review Rules
When reviewing maintenance output, look for:
- material work happened but the
.agents/ files were not reconciled
.agents/status.md contains guesses presented as facts.agents/tasks.md contains stale or non-actionable items.agents/record.md is missing stable references that future work will need.agents/journal.md is being used as current state instead of history- the five files disagree with each other
Validation Checklist
Do not claim maintenance success until you have checked:
- every affected
.agents/ file was considered
- verified facts and assumptions are clearly separated
- next actions are explicit
- stable references are captured
- historical notes were preserved when needed
- the
.agents/ files are consistent with the actual repo and latest verified state
This Skill Must Stop And Escalate When
- the material change is unclear
- the verified state is unknown
- the underlying domain skill has not been consulted
- the
.agents/ files conflict in a way that cannot be resolved from available evidence
Do not guess through contradictory project state.
