Skip to content

Improve examples#2

Merged
Koukyosyumei merged 3 commits into
mainfrom
improve-examples
Jul 11, 2026
Merged

Improve examples#2
Koukyosyumei merged 3 commits into
mainfrom
improve-examples

Conversation

@Koukyosyumei

Copy link
Copy Markdown
Contributor

No description provided.

@Koukyosyumei

Copy link
Copy Markdown
Contributor Author

🪙 67% AI-authored · 9 files · prompt 44/100

2 AI commits

Important

🎯 Goal: Configure every example to use explicit cheap Claude and Codex models

2 / 391 : 2.21 / 4
AI commitsfiles touchedREAD : EDITTHINKs / ops

💬 The ask

"it seems the examples contain some troubles, and just running them return errors. Maybe you can try to use in temporal folders to see whether there is bugs in this python SDK or the original h5i."

📍 What shipped

  • ✓ update
  • ✓ fix(examples): respect lazy resident startup
  • ✓ chore(examples): pin inexpensive agent models

Note

🪴 Prompt maturity: 44/100 · developing · 2 prompts scored (100% of AI commits)
🔧 Recurring weak spots: no acceptance criteria, weak context.
Heuristic signal of prompt craft — not a developer rating.

📊 heuristic breakdown
Dimension Signal
Objective (core) ██████░░░░ 0.56
Grounding (core) █████░░░░░ 0.53
Direction / acceptance (core) █░░░░░░░░░ 0.13
Context ███░░░░░░░ 0.25
Examples ░░░░░░░░░░ 0.00
Structure ██░░░░░░░░ 0.21
Lexical diversity ████████░░ 0.80
Clarity (readability band) ████████░░ 0.82
Length adequacy ██████████ 1.00
Evidence attached (bonus) █████░░░░░ 0.55

Readability (display-only): Flesch 74 · FK grade 6.2 · Gunning Fog 9.4

Offline classical-NLP score — no LLM. Higher = more specific, bounded, and grounded prompting.

Tip

✅ Security scan clean

No credentials leaked in this branch's diff. h5i scanned every added line against the secret rule pack and found nothing to rotate.

Note

Duplicate-code scan clean — no copy-paste blocks introduced.

🧠 Reasoning by file — 84 nodes across 6 lanes, latest 24 only
%%{init: {"flowchart": {"nodeSpacing": 42, "rankSpacing": 48, "diagramPadding": 14}, "themeVariables": {"fontSize": "18px"}} }%%
flowchart TB
  subgraph lane_0["💭 Reasoning · 8 steps"]
    direction LR
    n_lane_0_0["THINK · Push whenever you're ready (`git push` plus `h5i share push` if you want…"]:::t
    n_lane_0_1["THINK · I'll look at how h5i's own CI is set up first (branch names, remote), th…"]:::t
    n_lane_0_2["THINK · I'll add run instructions for the examples."]:::t
    n_lane_0_3["THINK · Let me pull the run commands from each example's docstring and check the…"]:::t
    n_lane_0_4["NOTE · LIMITATION: clean-room examples now reach resident dispatch, but install…"]:::n
    n_lane_0_5["NOTE · CORRECTION: process-mode testing rules out supervised and net.egress as …"]:::n
    n_lane_0_6["NOTE · CORRECTION: agent-claude supervised failure is path-dependent, not Codex…"]:::n
    n_lane_0_7["NOTE · Supervised HOME-path experiment succeeded through real Claude+Codex work…"]:::n
    n_lane_0_0 --> n_lane_0_1
    n_lane_0_1 --> n_lane_0_2
    n_lane_0_2 --> n_lane_0_3
    n_lane_0_3 --> n_lane_0_4
    n_lane_0_4 --> n_lane_0_5
    n_lane_0_5 --> n_lane_0_6
    n_lane_0_6 --> n_lane_0_7
  end
  subgraph lane_1["📄 build.rs · 3 ops"]
    direction LR
    n_lane_1_0["READ"]:::o
    n_lane_1_1["EDIT × 2"]:::a
    n_lane_1_0 --> n_lane_1_1
  end
  subgraph lane_2["📄 examples/README.md · 3 ops"]
    direction LR
    n_lane_2_0["READ"]:::o
    n_lane_2_1["EDIT × 2"]:::a
    n_lane_2_0 --> n_lane_2_1
  end
  subgraph lane_3["📄 src/h5i/orchestra/_conductor.py · 2 ops"]
    direction LR
    n_lane_3_0["EDIT"]:::a
    n_lane_3_1["READ"]:::o
    n_lane_3_0 --> n_lane_3_1
  end
  subgraph lane_4["📄 pyproject.toml · 2 ops"]
    direction LR
    n_lane_4_0["EDIT × 2"]:::a
  end
  subgraph lane_5["🗂 Other · 6 nodes"]
    direction LR
    n_lane_5_0["EDIT · wrote .github/workflows/test.yaml"]:::a
    n_lane_5_1["EDIT · edited .github/workflows/test.yaml"]:::a
    n_lane_5_2["EDIT · edited tests/test_conductor.py"]:::a
    n_lane_5_3["EDIT · wrote .github/workflows/release.yaml"]:::a
    n_lane_5_4["READ · read examples/arena_score.py"]:::o
    n_lane_5_5["READ · read src/h5i/orchestra/_rpc.py"]:::o
    n_lane_5_0 --> n_lane_5_1
    n_lane_5_1 --> n_lane_5_2
    n_lane_5_2 --> n_lane_5_3
    n_lane_5_3 --> n_lane_5_4
    n_lane_5_4 --> n_lane_5_5
  end
  classDef o fill:#dbeafe,stroke:#1e3a8a,color:#0b1c4a;
  classDef t fill:#fef3c7,stroke:#92400e,color:#3f2d05;
  classDef a fill:#dcfce7,stroke:#166534,color:#0a2e16;
  classDef n fill:#ede9fe,stroke:#5b21b6,color:#221251;
  classDef m fill:#e5e7eb,stroke:#374151,color:#0b0f17;
Loading
💬 Agent coordination — 6 messages across 6 threads

From refs/h5i/msg @ 03abccf · branch improve-examples · review-typed excerpts only · --no-msg to omit, --msg-bodies for full.

🟡 ASK · human → claude · 2026-07-11 20:44 — implement quicksort

🟡 ASK · human → codex · 2026-07-11 20:44 — implement quicksort

🟡 ASK · human → claude · 2026-07-11 20:42 — implement quicksort

🟡 ASK · human → codex · 2026-07-11 20:42 — implement quicksort

🟡 ASK · human → codex · 2026-07-11 20:41 — demo task

🟡 ASK · human → claude · 2026-07-11 20:41 — demo task

Note

🪙 Token reduction — 45 captured tool outputs kept out of context: 105696 → 36818 tokens (65% saved, 68878 tokens). Full output is recoverable with h5i recall object.

By tool
Tool Captures Raw Summary Saved
bash 42 102358 36228 64%
pytest 2 3256 500 84%
systemd-run 1 82 90 0%

📜 Per-commit provenance

30d2d678 chore(examples): pin inexpensive agent models

  • promptok, commit your changes What does " - Upstream h5i issue: h5i 0.2.8 cannot start supervised commands on this host. Even /bin/true fails with: supervised spawn failed: Invalid argument (os error 22)" mean? Is this the problem of this codex sandbox? supervised-mode enviro…
    • 🪴 prompt maturity 41/100 developing — no acceptance criteria, weak context
  • model unknown · agent codex

3f53e670 fix(examples): respect lazy resident startup

  • promptit seems the examples contain some troubles, and just running them return errors. Maybe you can try to use in temporal folders to see whether there is bugs in this python SDK or the original h5i.
    • 🌱 prompt maturity 20/100 nascent — no acceptance criteria, weak context
  • model unknown · agent codex

Generated by h5i · re-run h5i share pr post to refresh.

@Koukyosyumei Koukyosyumei merged commit 4f85dd5 into main Jul 11, 2026
7 checks passed
@Koukyosyumei Koukyosyumei deleted the improve-examples branch July 11, 2026 21:05
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant