Skip to content

docs/guides: refresh to ensemble narrative; clarify agent-run vs human-run commands#230

Merged
Koukyosyumei merged 12 commits into
mainfrom
improve-doc
Jun 26, 2026
Merged

docs/guides: refresh to ensemble narrative; clarify agent-run vs human-run commands#230
Koukyosyumei merged 12 commits into
mainfrom
improve-doc

Conversation

@Koukyosyumei

Copy link
Copy Markdown
Collaborator

No description provided.

…n-run commands

- guides index: drop features Overview card + JSON-LD entry, rework workflows card to ensemble framing
- prompt-injection: remove unnecessary 'h5i hook setup' code line (traces captured automatically)
- ai-code-provenance: reframe commit as agent-run (h5i capture commit replaces git commit; prompt/model/agent auto-captured), drop deprecated --prompt
- codex-collaboration: human runs only 'h5i msg setup' then asks agent in plain language; replace agent-run msg terminals with embedded i5h-protocol.svg
- workflows: rewrite body to follow README 60-second flow (install -> setup -> track -> sandbox -> ensemble -> recall) + update meta/og/JSON-LD
@Koukyosyumei Koukyosyumei changed the title Improve doc docs/guides: refresh to ensemble narrative; clarify agent-run vs human-run commands Jun 26, 2026
@Koukyosyumei

Copy link
Copy Markdown
Collaborator Author

🪙 8% AI-authored · 8 files · prompt 53/100

1 AI commit

Important

🎯 Goal: improve docs/guides per user feedback: fix stale narrative titles/summaries, clarify agent-run vs human-run commands, replace unneeded code blocks with protocol diagram, drop the features page, refactor workflows page to follow README flow

1 / 1281 : 1.2
AI commitsfiles touchedREAD : EDIT

💬 The ask

"Can you review https://github.com/h5i-dev/h5i/pull/227 Let's improve ./docs/guides. For example, I think: "the title and summary is old narative ### Detecting prompt injection in AI agent traces `$ h5i hook setup # capture OBSERVE/…"

📍 What shipped

  • ✓ update
  • ✓ update
  • ✓ update
  • ✓ swap the orders
  • ✓ docs/guides: refresh to ensemble narrative; clarify agent-run vs human-run commands

Note

🌿 Prompt maturity: 53/100 · proficient · 1 prompt scored (100% of AI commits)
🔧 Recurring weak spots: weak context.
Heuristic signal of prompt craft — not a developer rating.

📊 heuristic breakdown
Dimension Signal
Specificity ███████░░░ 0.73
Control / acceptance ████░░░░░░ 0.40
Context grounding █░░░░░░░░░ 0.13
Structure █████░░░░░ 0.45
Lexical diversity ████████░░ 0.79
Clarity (readability band) ██████████ 1.00
Length adequacy ██████████ 1.00

Readability (display-only): Flesch 55 · FK grade 10.4 · Gunning Fog 13.0

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 — 10016 nodes across 5 lanes, latest 24 only
%%{init: {"flowchart": {"nodeSpacing": 42, "rankSpacing": 48, "diagramPadding": 14}, "themeVariables": {"fontSize": "18px"}} }%%
flowchart TB
  subgraph lane_0["📄 docs/workflows/index.html · 7 ops"]
    direction LR
    n_lane_0_0["READ"]:::o
    n_lane_0_1["EDIT × 2"]:::a
    n_lane_0_2["READ"]:::o
    n_lane_0_3["EDIT × 3"]:::a
    n_lane_0_0 --> n_lane_0_1
    n_lane_0_1 --> n_lane_0_2
    n_lane_0_2 --> n_lane_0_3
  end
  subgraph lane_1["📄 docs/guides/index.html · 5 ops"]
    direction LR
    n_lane_1_0["READ"]:::o
    n_lane_1_1["EDIT × 4"]:::a
    n_lane_1_0 --> n_lane_1_1
  end
  subgraph lane_2["📄 docs/guides/ai-code-provenance/index.html · 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["📄 docs/guides/codex-claude-code-collaboration/index.html · 3 ops"]
    direction LR
    n_lane_3_0["READ"]:::o
    n_lane_3_1["EDIT"]:::a
    n_lane_3_2["READ"]:::o
    n_lane_3_0 --> n_lane_3_1
    n_lane_3_1 --> n_lane_3_2
  end
  subgraph lane_4["🗂 Other · 6 nodes"]
    direction LR
    n_lane_4_0["READ · read src/sandbox_policy.rs"]:::o
    n_lane_4_1["READ · read src/sandbox_policy.rs"]:::o
    n_lane_4_2["READ · read docs/guides/prompt-injection-detection-for-agents/index.html"]:::o
    n_lane_4_3["EDIT · edited docs/guides/prompt-injection-detection-for-agents/index.html"]:::a
    n_lane_4_4["READ · read README.md"]:::o
    n_lane_4_5["READ · read docs/_static/i5h-protocol.svg"]:::o
    n_lane_4_0 --> n_lane_4_1
    n_lane_4_1 --> n_lane_4_2
    n_lane_4_2 --> n_lane_4_3
    n_lane_4_3 --> n_lane_4_4
    n_lane_4_4 --> n_lane_4_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

📜 Per-commit provenance

2363ba45 docs/guides: refresh to ensemble narrative; clarify agent-run vs human-run commands

  • promptCan you review https://github.com/h5i-dev/h5i/pull/227 Let's improve ./docs/guides. For example, I think: "the title and summary is old narative ### Detecting prompt injection in AI agent traces $ h5i hook setup # capture OBSERVE/THINK/ACT traces per session is not n e…
    • 🌿 prompt maturity 53/100 proficient — weak context
  • model claude-opus-4-8 · agent claude-code

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

@Koukyosyumei Koukyosyumei merged commit 6d467d5 into main Jun 26, 2026
5 checks passed
@Koukyosyumei Koukyosyumei deleted the improve-doc branch June 26, 2026 15:32
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