Session Status

Inspect bundle runtime state for: $ARGUMENTS

Overview

Implementation status: code-backed
Local entry script: <bundle-root>/session-status/run.py
Primary purpose: expose the bundle root, config path, output root, and recent command history so the agent can continue the workflow from explicit context instead of guesswork
Research layer: operational monitoring (Cross-cutting concern for all workflow stages)
Workflow role: operational control surface spanning artifact reuse, history lookup, and monitoring handoff
Local executor guarantee: return the current runtime root, config path, last symbol, Feishu mode, and recent history for the chosen execution context

Use When

The user asks what the bundle has been doing.
The user wants to know where runtime output is being written.
The user wants to confirm the current Feishu mode or the last-used symbol.
The agent needs to decide whether to rerun, reuse artifacts, or continue a monitoring workflow.
The user wants to understand current session state and context.
The user wants to resume work from a previous session.
The user wants to audit recent activity for troubleshooting.

Do Not Use When

The user wants analysis or decision output. Use a market-facing skill.
The user wants to inspect a specific prior report in detail. Use reports or analysis-history.
The user wants a data-freshness diagnosis rather than runtime context. Use data-sync.
The user wants portfolio performance tracking. Use paper-trading.

Inputs

No required positional arguments.
Optional --context through the local script if the caller needs a non-default execution context.

Execution

Step 1: Gather runtime context

Collect comprehensive session state information:

Runtime configuration:

Bundle root directory path
Config file path and contents
Output directory path
Execution context (default or custom)
Environment variables (if relevant)
Python version and dependencies

Session state:

Integration status:

Feishu mode (enabled/disabled)
Notification settings
External integrations (if any)
API connections (if any)

Recent activity:

Step 2: Inspect artifact state

Audit available artifacts and their status:

Artifact inventory:

Analysis reports (count, latest date)
Stock data packets (count, latest date)
Backtest results (count, latest date)
Paper trading records (count, latest date)
Screening results (count, latest date)
Dashboard snapshots (count, latest date)

Artifact freshness:

Latest artifact timestamp
Stale artifacts (> 7 days old)
Orphaned artifacts (no longer referenced)
Incomplete artifacts (partial writes)

Artifact locations:

Step 3: Assess workflow state

Determine where the workflow left off:

Workflow stage identification:

Stage 1 (Universe Formation): Last watchlist import or screening
Stage 2 (Data Collection): Last data sync or market-data run
Stage 3 (Data Cleaning): Last normalization or validation
Stage 4 (Feature Engineering): Last market-analyze or technical-scan
Stage 5 (Backtesting): Last backtest-evaluator run
Stage 6 (Risk Management): Last decision-support or strategy-design
Stage 7 (Live Trading): Last paper-trading execution

Workflow continuity:

Can workflow resume from last state?
Are artifacts reusable or stale?
Are dependencies satisfied?
Are there blocking issues?

Next logical steps:

Based on last activity, what should happen next?
Which artifacts can be reused?
Which skills should be invoked?
What validation is needed before proceeding?

Step 4: Diagnose operational issues

Identify any operational problems:

Common issues:

Stale artifacts: Artifacts too old for current use
Missing dependencies: Required data or artifacts not available
Configuration errors: Invalid or missing config settings
Permission issues: Cannot write to output directories
Disk space issues: Output directory full or near capacity
Integration failures: Feishu or other integrations not working

For each issue:

Issue type and severity
Impact on workflow
Recommended resolution
Workaround (if available)

Step 5: Generate status report

Organize findings into structured report:

Part 1: Runtime Configuration

Bundle root directory
Config file path
Output directory path
Execution context
Integration status (Feishu, etc.)

Part 2: Session State

Session start time and duration
Last symbol used
Last skill executed
Last execution timestamp
Active watchlists

Part 3: Recent Activity

Recent commands (last 10-20 with timestamps)
Recent symbols analyzed
Recent artifacts created
Recent errors or warnings
Recent data syncs

Part 4: Artifact Inventory

Analysis reports (count, latest)
Stock data packets (count, latest)
Backtest results (count, latest)
Paper trading records (count, latest)
Screening results (count, latest)
Dashboard snapshots (count, latest)

Part 5: Artifact Freshness

Latest artifact timestamp
Stale artifacts (count and list)
Orphaned artifacts (count and list)
Incomplete artifacts (count and list)

Part 6: Workflow State

Current workflow stage
Last completed stage
Next logical steps
Artifacts available for reuse
Dependencies satisfied/missing

Part 7: Operational Issues

Issues detected (count and list)
Severity distribution
Impact on workflow
Recommended resolutions

Part 8: Recommendations

Immediate actions (critical issues)
Artifact cleanup (stale/orphaned)
Workflow resumption (next steps)
Monitoring setup (ongoing)

Step 6: Run the local executor

python3 <bundle-root>/session-status/run.py

Step 7: Use result as operational audit

When delivering results, focus on actionable context:

Operational context interpretation:

This is runtime state, not analysis or recommendations
Artifact inventory is for reuse decisions, not performance assessment
Recent activity is for continuity, not evaluation
Workflow state is for resumption, not validation

Artifact reuse guidance:

State which artifacts are fresh enough to reuse
State which artifacts are stale and should be regenerated
State which artifacts are missing and need generation
State dependencies between artifacts

Workflow resumption guidance:

State where workflow left off
State next logical steps
State which skills to invoke next
State validation needed before proceeding

Issue resolution guidance:

State operational issues clearly
Provide specific resolution steps
Prioritize by severity and impact
Offer workarounds when available

Output Contract

Minimum local executor output: human-readable text beginning with A-Stockit runtime status.
Core fields: bundle root, config file, output directory, Feishu mode, last symbol, and recent context history.
Side effects: adds one session-status history record for the current execution context.
Caller-facing delivery standard:
- Eight-part structure: Runtime configuration, session state, recent activity, artifact inventory, artifact freshness, workflow state, operational issues, recommendations
- Operational context only: Present as runtime state, not analysis or recommendations
- Artifact reuse guidance: State which artifacts are fresh/stale/missing and reusable
- Workflow resumption guidance: Identify likely next useful artifact or skill for resuming work
- Issue diagnosis: Specific operational issues with severity and resolution steps
- Scope clarity: This is a context surface, not a semantic summary of every run
- Actionable recommendations: Immediate actions, cleanup, resumption steps, monitoring
- No analysis claims: Runtime state only, no market views or investment recommendations

Failure Handling

Parse and argument errors: non-zero exit with a readable 命令错误 message.
Unexpected runtime issues: readable failure text beginning with 执行失败:.
Missing config or output directories: report issue and recommend initialization.
Corrupted history or artifacts: flag for cleanup and regeneration.
Permission issues: report specific permission errors and resolution steps.

Key Rules

Use this skill for operational context only.
Prefer it before rerunning workflows just to locate prior outputs.
Treat it as the staging surface for artifact reuse, monitoring continuation, and workflow resumption.
Artifact inventory is for reuse decisions, not performance evaluation.
Recent activity is for continuity, not retrospective analysis.
Workflow state is for resumption, not validation.
Issue diagnosis must be specific and actionable.
Recommendations must be prioritized by severity and impact.
No analysis or investment claims. This is operational monitoring only.

Composition

Pairs naturally with reports, analysis-history, data-sync, and paper-trading.
Should be used before artifact-dependent skills to check reusability.
Can inform data-sync about what needs refreshing.
Can inform reports and analysis-history about available artifacts.
Can inform workflow resumption decisions across all skills.