docs: add Harbor+Daytona instructions to all agent guides and navigation

Rewrite docs/DAYTONA.md to position Harbor+Daytona (--environment-type
daytona) as the recommended approach for production benchmark runs with
full artifact collection. Add Daytona routing to root and local
CLAUDE.md/AGENTS.md files, START_HERE_BY_TASK.md, and task_routes.json.
Fix pre-existing issues: remove stale tasks/ directory references from
sync_agent_guides.py and docs_consistency_check.py, remove dangling
RESULT_DIRECTORY_SPEC.md reference, regenerate script registry and index.

Author: sjarmak

AGENTS.md

@@ -16,6 +16,7 @@ full operations manual.
 
 ## Fast Routing By Intent
 - Launch or rerun benchmarks: `docs/START_HERE_BY_TASK.md` -> "Launch / Rerun Benchmarks"
+- Run benchmarks on Daytona (cloud, no Docker needed): `docs/DAYTONA.md`
 - Monitor / status: `docs/START_HERE_BY_TASK.md` -> "Monitor Active Runs"
 - Triage failures: `docs/START_HERE_BY_TASK.md` -> "Triage Failed Tasks"
 - Compare configs / MCP impact / IR: `docs/START_HERE_BY_TASK.md` -> "Analyze Results"
@@ -25,7 +26,6 @@ full operations manual.
 ## Local Guides
 - `scripts/AGENTS.md` - script categories, safe usage, one-off handling
 - `configs/AGENTS.md` - run launcher wrappers and confirmation gate policy
-- `tasks/AGENTS.md` - task metadata and validation workflow
 - `docs/AGENTS.md` - documentation IA and canonical vs archive guidance
 
 ## Compaction / Handoff Checkpoints

CLAUDE.md

@@ -16,6 +16,7 @@ full operations manual.
 
 ## Fast Routing By Intent
 - Launch or rerun benchmarks: `docs/START_HERE_BY_TASK.md` -> "Launch / Rerun Benchmarks"
+- Run benchmarks on Daytona (cloud, no Docker needed): `docs/DAYTONA.md`
 - Monitor / status: `docs/START_HERE_BY_TASK.md` -> "Monitor Active Runs"
 - Triage failures: `docs/START_HERE_BY_TASK.md` -> "Triage Failed Tasks"
 - Compare configs / MCP impact / IR: `docs/START_HERE_BY_TASK.md` -> "Analyze Results"
@@ -25,7 +26,6 @@ full operations manual.
 ## Local Guides
 - `scripts/AGENTS.md` - script categories, safe usage, one-off handling
 - `configs/AGENTS.md` - run launcher wrappers and confirmation gate policy
-- `tasks/AGENTS.md` - task metadata and validation workflow
 - `docs/AGENTS.md` - documentation IA and canonical vs archive guidance
 
 ## Compaction / Handoff Checkpoints

configs/AGENTS.md

@@ -11,6 +11,7 @@ Use this file when working in `configs/` launchers and run orchestration wrapper
 - Start with `configs/_common.sh` for shared run policy and confirmation behavior.
 - Use `configs/run_selected_tasks.sh` for selected-task execution flows.
 - Use `configs/*_2config.sh` wrappers for paired baseline/MCP runs.
+- For Daytona cloud execution, add `--environment-type daytona` to `harbor run` commands. See `docs/DAYTONA.md` for prerequisites and capacity planning.
 
 ## When Editing
 - Preserve `confirm_launch()` gating behavior.

configs/CLAUDE.md

@@ -11,6 +11,7 @@ Use this file when working in `configs/` launchers and run orchestration wrapper
 - Start with `configs/_common.sh` for shared run policy and confirmation behavior.
 - Use `configs/run_selected_tasks.sh` for selected-task execution flows.
 - Use `configs/*_2config.sh` wrappers for paired baseline/MCP runs.
+- For Daytona cloud execution, add `--environment-type daytona` to `harbor run` commands. See `docs/DAYTONA.md` for prerequisites and capacity planning.
 
 ## When Editing
 - Preserve `confirm_launch()` gating behavior.