Feat/add controller performance cache

[cardosofede]

No description provided.

[rapcmia]

Commit 863d6e5

• Fix/initial positions failure hummingbot#8240 - build local hbot lib ✅
• Feat/add controller performance cache #163 - build local docker using hbot lib ✅
• Both integrated with this Feat / general upgrade and GitHub actions condor#102 ✅

Test condor/102

• Setup a trading agent on binance perpetual using AVAX-USDT ✅
• Test Executors page active executor data ✅
  • Started one grid executor
  • Executors page showed Active(1)
    • The active row displayed age, type, connector, pair, side, volume, and PnL
    • Clicking the executor opened the detail panel with grid details, chart, PnL, and volume
    • The active executor showed a stop button
• Test Executors page performance values and summaries ✅
  • The grid executor volume matched the page summary
  • The PnL matched the page summary
  • Active performance values and summaries displayed correctly for the running executor
• Test Executors page archived/completed executor visibility if that path is used ✅
  • The executor is terminated, it moves into History (1) after completion
  • close type was added for the archived executor
  • The archived row kept the last recorded values for PnL, volume, and age
  • Opening the terminated executor showed the same details and they matched the row data
• Test Trade or positions summary for the same perpetual controller ✅
  • Successfully started a bot using grid_strike on docker image PR8240
  • Set keep_position: true and observed when executor is terminated
    • Check trade page, positions tab and confirmed position held is displayed
    • Tested again with grid executor and call executor/stop from hummingbot-api still works ok
• Test Bots page if it reads controller performance affected by the API cache change
  • PNL does not match ❌
    
    • Also found an error when i check the deployed bot from condor i got error:

      2026-05-19 14:39:29,992 - 17 - hummingbot.core.utils.async_utils - ERROR - Unhandled error in background task: object of type 'int' has no len()
      Traceback (most recent call last):
        File "/home/hummingbot/hummingbot/core/utils/async_utils.py", line 9, in safe_wrapper
          return await c
                 ^^^^^^^
        File "/home/hummingbot/hummingbot/client/command/status_command.py", line 137, in status_check_all
          self.notify(await self.strategy_status())
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "/home/hummingbot/hummingbot/client/command/status_command.py", line 78, in strategy_status
          st_status = self.trading_core.strategy.format_status()
        File "/home/hummingbot/hummingbot/strategy/strategy_v2_base.py", line 506, in format_status
          lines.extend(controller.to_format_status())
                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
        File "/home/hummingbot/controllers/generic/grid_strike.py", line 153, in to_format_status
          f"NOT_ACTIVE: {len(level.custom_info['levels_by_state'].get('NOT_ACTIVE', []))}",
                         ~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
      TypeError: object of type 'int' has no len()
      

      • Check standalone hummingbot/8240 and development branch (source and docker) using same controller v2 (generic.gridstrike), not able to reproduce after use status command ✅
• Test archived bots view if you use it ✅
• Test one stop/start lifecycle flow for the affected controller ✅
  • When started, deployed v2 is displayed as active on bots page
  • When stopped, exits the bots page and goes to archive
    • Most of the test had orders filled perhaps these are recorded successfully

---

Deploy using hummingbot/8240

• This test is still pending ❗
• Test starting a perpetual controller/strategy that depends on initial position handling
• Test that the initial position is detected or initialized correctly on startup
• Test that position mode is set correctly for the affected perpetual pair
• Test that the controller keeps running normally after startup without position-related failure
• Also @nikspz can you confirm these checklist testing the hummingbot standalone PR too, thanks 🙏🏼

[rapcmia]

Commit 4173791

Also found an error when i check the deployed bot from condor i got error:

• Run status command works

---

Test with hummingbot/8240

• Build local docker image for Fix/initial positions failure hummingbot#8240
• Deploy generic.pmm_mister using condor/102 and this PR
• Test starting a perpetual controller/strategy that depends on initial position handling ✅
  
  • Started a perpetual V2 controller with initial_positions configured.

    "amount": "2.7",
    "connector_name": "hyperliquid_perpetual",
    "side": "BUY",
    "trading_pair": "HYPE-USD"
    

  • Confirmed startup completed and the controller created normal open orders.
  • Confirmed the orchestrator logged Created initial position for controller ..., showing the initial position was loaded successfully.
    
  • Checking the log files, confirmed no unexpected startup failure occurred while handling the initial position
  • Confirmed no live exchange position was created automatically, which matches the intended design of initial_positions
• Test that the initial position is detected or initialized correctly on startup ✅
  • The startup log also confirmed the initial position was created, showing the configured value was initialized correctly and exposed in controller bot page.
  • The generic.pmm_mister adjusted its target/min/max base logic from the seeded initial position shown in status.
• Test that position mode is set correctly for the affected perpetual pair ✅
• Test that the controller keeps running normally after startup without position-related failure ✅