diff options
| author | main <main@swarm.moe> | 2026-03-19 10:17:07 -0400 |
|---|---|---|
| committer | main <main@swarm.moe> | 2026-03-19 10:17:07 -0400 |
| commit | 08a1139eaa7a4862ab8c0e5fb5fc6845fc711208 (patch) | |
| tree | ded498d30e1d84c17b3e6dbf80594b5b62faa804 /assets/codex-skills/mcp-bootstrap/references/checklist.md | |
| download | libmcp-08a1139eaa7a4862ab8c0e5fb5fc6845fc711208.zip | |
Initial libmcp 1.0.0
Diffstat (limited to 'assets/codex-skills/mcp-bootstrap/references/checklist.md')
| -rw-r--r-- | assets/codex-skills/mcp-bootstrap/references/checklist.md | 15 |
1 files changed, 15 insertions, 0 deletions
diff --git a/assets/codex-skills/mcp-bootstrap/references/checklist.md b/assets/codex-skills/mcp-bootstrap/references/checklist.md new file mode 100644 index 0000000..b20a836 --- /dev/null +++ b/assets/codex-skills/mcp-bootstrap/references/checklist.md @@ -0,0 +1,15 @@ +# Checklist + +Use this checklist when reviewing a `libmcp` consumer. + +- Does a stable host own the public session? +- Is worker fragility isolated behind an explicit replay policy? +- Are replay contracts typed and local to the request surface? +- Are faults typed and connected to recovery semantics? +- Do nontrivial tools default to porcelain output? +- Is structured JSON still available where exact consumers need it? +- Are inputs normalized where the semantics are still unambiguous? +- Are health and telemetry available? +- Is event telemetry append-only and useful for postmortem analysis? +- Does the recovery test matrix cover the failure modes actually observed? +- Is the installed `$mcp-bootstrap` skill sourced from this repository? |