fix(compression): make manual compression feedback truthful

[bobashopcashier]

What does this PR do?

Makes manual compression feedback truthful across both user-facing entrypoints: gateway /compress and CLI manual compression. No-op results are detected by comparing the pre/post transcript, token counts are labeled as rough transcript estimates, and cases where fewer messages still yield a higher rough estimate get an explicit explanatory note instead of misleading success copy.

Related Issue

Fixes #6202
Fixes #6217
Related: #6270
Separate env/pairing hardening is tracked in PR #6956

Type of Change

• 🐛 Bug fix (non-breaking change that fixes an issue)
• ✨ New feature (non-breaking change that adds functionality)
• 🔒 Security fix
• 📝 Documentation update
• ✅ Tests (adding or improving test coverage)
• ♻️ Refactor (no behavior change)
• 🎯 New skill (bundled or hub)

Changes Made

• added a shared manual-compression feedback formatter so gateway and CLI use the same no-op detection and estimate wording
• updated gateway/run.py to report no-op compression truthfully and explain rough transcript estimate increases after summary rewrites
• updated cli.py so manual compression uses the same truthful messaging instead of always printing a success banner
• added focused regression coverage for gateway /compress and CLI manual compression output

How to Test

1. Run /Users/kennyxie/Documents/Code/clawspace/hermes-agent/.venv/bin/python -m pytest tests/gateway/test_compress_command.py tests/cli/test_manual_compress.py -q
2. Trigger manual compression on an unchanged transcript and verify Hermes reports a no-op rather than a successful compression
3. Trigger manual compression on a transcript that compresses to fewer messages but a higher rough estimate and verify Hermes explains why

Checklist

Code

• I've read the Contributing Guide
• My commit messages follow Conventional Commits (fix(scope):, feat(scope):, etc.)
• I searched for existing PRs to make sure this isn't a duplicate
• My PR contains only changes related to this fix/feature (no unrelated commits)
• I've run pytest tests/ -q and all tests pass
• I've added tests for my changes (required for bug fixes, strongly encouraged for features)
• I've tested on my platform: macOS 26.4

Documentation & Housekeeping

• I've updated relevant documentation (README, docs/, docstrings) — or N/A
• I've updated cli-config.yaml.example if I added/changed config keys — or N/A
• I've updated CONTRIBUTING.md or AGENTS.md if I changed architecture or workflows — or N/A
• I've considered cross-platform impact (Windows, macOS) per the compatibility guide — or N/A
• I've updated tool descriptions/schemas if I changed tool behavior — or N/A

Screenshots / Logs

• /Users/kennyxie/Documents/Code/clawspace/hermes-agent/.venv/bin/python -m pytest tests/gateway/test_compress_command.py tests/cli/test_manual_compress.py -q → 4 passed

[teknium1]

Merged via #7541 with authorship preserved. Thanks for the contribution!