Skip to content

SKILL.md

Latest commit

 

History

History
109 lines (83 loc) · 4.79 KB

SKILL.md

File metadata and controls

109 lines (83 loc) · 4.79 KB
Raw
name explain-lading-config
description Explains a lading.yaml config file from the regression test suite, using the lading Rust source as ground truth for field meanings and defaults.
user_invocable true
argument-hint [experiment name]
model sonnet

explain-lading-config

Explain what a lading regression test config does, grounded in lading source code.

Quick Start

# 1. Verify the lading checkout exists and is on a known branch
bash .agents/skills/explain-lading-config/scripts/validate-lading-checkout.sh

# 2. Resolve $ARGUMENTS to a lading.yaml path (exact/substring/glob/path)
bash .agents/skills/explain-lading-config/scripts/resolve-lading-config.sh "$ARGUMENTS"

# 3. Read the resolved file, then ground every field in lading source
#    (see references/source-reading.md for the full strategy).

# 4. Write up the explanation following references/explanation-template.md.

Defaults must be resolved to concrete values, not function names. Full workflow below.

Step 1: Validate lading checkout

Run .agents/skills/explain-lading-config/scripts/validate-lading-checkout.sh.

  • Exit 0: script prints the current branch on stdout. If it is not main, warn the user that explanations are grounded in a non-main branch, then continue.
  • Exit non-zero: the script prints a suggested git clone command on stderr. Relay that to the user and stop.

Override the checkout location with LADING_DIR if needed.

Step 2: Determine target file

Use .agents/skills/explain-lading-config/scripts/resolve-lading-config.sh to avoid ad-hoc matching. The script enumerates experiments under any cases/ directory within test/regression/ (e.g. quality_gates/cases/, logs/<group>/cases/) and test/regression/x-disabled-cases/ (disabled). Each experiment is a <case>/lading/lading.yaml addressed by its case-directory name; disabled rows are flagged with a trailing (disabled) column in the listing. The ebpf/ subtrees (split-mode and config-only) are intentionally out of scope; if a user asks about one, tell them this skill doesn't cover it yet.

The script handles path-like inputs, substring case names, and shell globs (*, ?).

If $ARGUMENTS is provided: run resolve-lading-config.sh "$ARGUMENTS".

  • Exit 0: stdout is the resolved absolute path; read it.
  • Exit 3 (ambiguous): stderr lists candidates.
    • ≤ 4 candidates: use AskUserQuestion to pick one, then read that path.
    • > 4 candidates (a broad substring like i can match 20+): do not try to force them into AskUserQuestion. Print the experiment names as a short bulleted list and ask the user to narrow the query and re-invoke /explain-lading-config <name>.
  • Exit 2 (not found): stderr may include "did you mean?" suggestions — if present, offer the suggestions to the user via AskUserQuestion (up to 4 options) or as a short list; if not, relay the error and stop.
  • Exit 4 (wrong repo): the script is being run from outside the agent repo. Relay the error verbatim and stop — the user needs to cd into the repo.

If the resolved path contains /x-disabled-cases/, flag this explicitly in the explanation — the experiment exists on disk but is not currently executed by SMP. Otherwise a user may assume it's live.

Reading very large configs: multi-sender configs (e.g. uds_dogstatsd_20mb_12k_contexts_20_senders, ~870 lines) are usually block-copies of one template with a few fields varying (typically only seed). Before a full Read, check size and duplication:

wc -l <path>                                    # scale check
grep -c '^  - ' <path>                          # top-level list entries
yq '.generator | length' <path> 2>/dev/null     # if yq is present

For highly-duplicated configs, Read only the first block (plus the blackhole/target_metrics sections) and report the generator as "N identical copies, seed differs" instead of walking every block. Spot- check one later block to confirm uniformity.

If $ARGUMENTS is omitted: run resolve-lading-config.sh with no argument. It emits <experiment>\t<path> lines for every discovered config.

Print the experiment names as a plain bulleted list to the user (preserving the (disabled) markers) and ask them to type the name (or re-invoke the skill with /explain-lading-config <name>).

Step 3: Read the lading codebase for context

Before explaining, read the lading source files that ground the populated sections of the config. The detailed strategy (variant-to-module mapping, grep-before-Read invariants, fallback for renamed files) lives in references/source-reading.md — read it now.

Step 4: Explain the config

Write the explanation following the structure in references/explanation-template.md (generator summary, aggregate load, blackhole sinks, target metrics, source references). Read it now.