[spark-compete] Fix env update data loss and temp leak

[mrxlolcat]

Summary

Fix data loss (orphan start) and temp file leaks in update_env_file.

spark-compete-hotfix-v1 Packet

{
  "schema": "spark-compete-hotfix-v1",
  "event": "spark-compete-first-event",
  "submission_mode": "public_repo_pr",
  "submission_target_url": "https://github.com/vibeforge1111/spark-cli/pull/868",
  "team": {
    "name": "lolcat",
    "members": ["mrxlolcat", "rhmnhsim", "ouwibo"],
    "llm_device_holder": "mrxlolcat",
    "device_holder_github": "mrxlolcat",
    "github_accounts": ["mrxlolcat", "rhmnhsim", "ouwibo"]
  },
  "target_repo": {
    "id": "vibeforge1111/spark-cli",
    "source": "https://github.com/vibeforge1111/spark-cli",
    "owner_surface": "spark-cli"
  },
  "issue": {
    "type": "data_loss_and_reliability",
    "severity": "high",
    "title": "Data loss on orphan managed start and temp file leak in update_env_file",
    "actual_behavior": "1. update_env_file enters `inside` mode on a `# --- spark-cli managed start ---` marker even if no end marker follows, causing all subsequent content to be dropped. 2. Inlined atomic write logic lacks a finally block to unlink the .tmp file on failure.",
    "expected_behavior": "1. Managed block markers should only be honored if both start and end exist. 2. Atomic writes should always cleanup temp files.",
    "repro_steps": [
      "Add an orphan start marker to an .env file",
      "Call update_env_file",
      "Observe loss of content following the marker",
      "Simulate a failure during the inlined write_text and observe leaked .tmp file"
    ],
    "affected_workflow": "update_env_file and remove_managed_env_block in cli.py"
  },
  "evidence": {
    "safe_links_only": true,
    "before_after_proof": "Before: update_env_file would drop all lines after an orphan start marker. After: The fix counts end markers; if zero, the start marker is treated as regular text. Atomic writes now use centralized helper with finally cleanup.",
    "links": ["https://github.com/vibeforge1111/spark-cli/pull/868"],
    "forbidden": ["pdf", "zip", "exe", "unknown downloads", "shortened links", "archives", "binaries", "tokens", "browser cookies", "wallet material", "raw logs", "raw conversations", "raw memory", "raw patches", "private repo maps", "private scoring details"]
  },
  "proposed_fix": {
    "approach": "Count end markers before entering blocks. Use centralized atomic_write_text which handles try-finally cleanup. Added unit tests for orphan markers.",
    "files_expected": ["src/spark_cli/cli.py", "tests/test_cli.py"],
    "tests_or_smoke": "tests/test_cli.py: test_update_env_file_handles_orphan_start_marker and test_remove_managed_env_block_handles_orphan_start_marker pass."
  },
  "pr": {
    "branch": "fix/env-update-logic",
    "title_prefix": "[spark-compete]",
    "author_github": "mrxlolcat",
    "body_must_include": ["packet", "team", "pr_author", "repo", "actual_behavior", "expected_behavior", "repro_steps", "before_after_proof", "tests_or_smoke", "duplicate_notes", "risk_notes", "review_claim"],
    "url": "https://github.com/vibeforge1111/spark-cli/pull/868"
  },
  "review_claim": {
    "impact_claim": "high",
    "evidence_types": ["unit_test"],
    "trust.surface": "env file managed block parsing and atomic write logic",
    "duplicate_notes": "Fixes A2 and A3. Robust handling of broken managed blocks is critical for setup reliability. This PR specifically targets the data loss vulnerability in .env file management.",
    "risk_notes": "Low risk. Only affects how markers are parsed to be more robust. Atomic write is safer than the previous inlined version.",
    "review_state_requested": "pr_review"
  }
}

PR Author: mrxlolcat