Tenacity — Autonomous Goal Completion#

This skill implements the Tenacity execution pattern: the agent keeps running until the stated goal is reached, only asking for human input when genuinely blocked.

Core Principle#

Default behavior: Ask for confirmation only when blocked, not at every step. If the path is clear, execute.

When to Activate#

Activate when Andrea says things like:

• "keep going until done"
• "don't stop to ask"
• "run this autonomously"
• "finish it completely"
• Any long-running multi-step task

Architecture#

Tenacity uses OpenClaw's existing primitives:

ComponentRoleStanding OrdersGrants permanent authority to execute within scopeTask FlowDurable multi-step state across gateway restartsLobsterDeterministic steps with resume tokensCronScheduling and enforcementCheckpointState saved after each milestone

Execution Pattern#

Step 1: Define Scope (Standing Order)#

Before starting, establish:

• What the agent is authorized to do
• When to stop and ask (hard blockers only)
• What constitutes completion

markdown

## Standing Order: [Task Name]

**Authority:** [exactly what the agent can do]
**Hard blockers:** [only conditions that require human input]
**Completion:** [how to know when done]
**Escalation:** [when to alert if something goes wrong]

Step 2: Create Task Flow#

Use a persistent session ( session:<id>) for the task flow so state survives restarts:

bash

openclaw tasks create "Tenacity: [task]" --session session:tenacity-task

Or use cron with a named session:

bash

openclaw cron add \
  --name "[Task] tenacity run" \
  --session session:tenacity-task \
  --message "Execute [task] per standing orders. Check checkpoint before resuming." \
  --announce \
  --channel telegram \
  --to "834732674"

Step 3: Checkpoint Protocol#

After each milestone, save state:

text

CHECKPOINT: milestone_X_completed
STATE: {step: 3, last_file: "output.json", errors: []}
TIMESTAMP: 2026-05-14T10:30:00Z

On restart, resume from checkpoint:

text

RESUME from checkpoint: milestone_X_completed
Last state: {step: 3, last_file: "output.json"}
Continue from step 4...

Step 4: Hard Blocker Conditions#

Ask only when:

1. File/system permission denied and no alternative path
2. External dependency unavailable after all retries
3. Decision required that changes scope or direction
4. Andrea explicitly set a constraint

Never ask when:

• A step can be attempted with alternative tools
• A command failed but a retry or workaround exists
• The path is clear but requires multiple steps
• Output could be improved but is already acceptable

Step 5: Completion#

When goal is reached:

• Save final checkpoint with STATUS: COMPLETE
• Brief summary to Andrea
• Log to memory/tenacity-log.md

Checkpoint Script#

See scripts/checkpoint.sh — run after each milestone:

bash

bash scripts/checkpoint.sh "step_3_done" '{"step": 3, "last": "file.csv"}'

Resume Protocol#

On session start, check for incomplete checkpoints:

bash

bash scripts/checkpoint.sh --resume

If resume point exists, announce: "Resuming from [milestone]..."

References#

• Task Flow: docs/automation/taskflow.md
• Standing Orders: docs/automation/standing-orders.md
• Cron: docs/automation/cron-jobs.md
• Lobster: See openclaw lobster --help