investigate

Name: investigate
Author: Borda

23stars3forks

How to install investigate?

borda/ai-rig/investigate

$npx -y skills add Borda/AI-Rig --skill investigate

Installs into the current project.

Summary

Systematic diagnosis for unknown failures — local environment, tool setup, CI vs local divergence, hook misbehavior, and runtime anomalies. Gathers signals broadly, ranks hypotheses, uses adversarial review (Codex or foundry:challenger) for ambiguous cases, probes each, and repor

Files · 1

View on GitHub

SKILL.md

1	<objective>
2
3	Diagnose unknown failures: broken local setup, environment mismatch, tool misbehavior, hook problems, CI vs local divergence, permission errors, runtime anomalies. Gather signals broadly, eliminate hypotheses systematically, report confirmed root cause + recommended next skill. No fixes — diagnosis only.
4	NOT for: known Python test failures with traceback (use `/develop:debug` (requires `develop` plugin)); `.claude/` config quality sweep (use `/foundry:audit`).
5
6	</objective>
7
8	<inputs>
9
10	- $ARGUMENTS: required — symptom, question, or failing command, e.g.:
11	- `"hooks not firing on Save"`
12	- `"codex:codex-rescue agent exits 127 on this machine"`
13	- `"/calibrate times out every run"`
14	- `"CI fails but passes locally"`
15	- `"uv run pytest can't find conftest.py"`
16
17	- `--fast`: optional flag — skip Step 4 adversarial Codex review; use when speed matters more than thoroughness or Codex unavailable.
18
19	If $ARGUMENTS empty or too vague, use AskUserQuestion: "What exactly is failing or behaving unexpectedly? Include the command and any error output you can share."
20
21	</inputs>
22
23	<compaction>
24	Key boundary 1: end of Step 2 (signals.md written to run-dir), before Step 3 rank hypotheses.
25	Key boundary 2: end of Step 3 (hypotheses.md written), refreshed again after each Step 5 probe verdict — so a mid-loop compaction does NOT re-rank or re-probe.
26	Preserve: INVESTIGATE_RUN, symptom.txt, signals.md, hypotheses.md paths; adversarial-review path (codex/challenger) if Step 4 ran; probe ledger (hypothesis → Confirmed/Ruled-out/Inconclusive).
27	Terminal path: end of Step 6 (report + follow-up gate complete).
28	</compaction>
29
30	<workflow>
31
32	Task hygiene: load and follow the protocol below.
33	```bash
34	# loads: compaction-contract.md
35	# audit-skip: resilience-replication
36	_FS=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/resolve_shared_path.py" foundry skills/_shared 2>/dev/null \|\| echo "plugins/cc_foundry/skills/_shared") # timeout: 5000
37	cat "$_FS/task-hygiene.md"
38	```
39
40	Task tracking: TaskCreate tasks for Gather, Hypothesise, Probe, Report; mark in_progress/completed as you go.
41
42	## Step 1: Parse symptom and scope
43
44	```bash
45	export CSID="${CLAUDE_CODE_SESSION_ID:-$PPID}"
46	KEEP_ITEMS=""
47	if [[ "$ARGUMENTS" =~ --keep[[:space:]]\"([^\"]+)\" ]]; then
48	KEEP_ITEMS="${BASH_REMATCH[1]}"
49	fi
50	ARGUMENTS=$(echo "$ARGUMENTS" \| sed 's/--keep "[^"]*"//g')
51	rm -f .temp/state/skill-contract.md ${TMPDIR:-/tmp}/investigate-verdicts-${CSID} # clear stale contract + probe ledger (compaction-contract.md §Lifecycle) # timeout: 5000
52	echo "$KEEP_ITEMS" > "${TMPDIR:-/tmp}/investigate-keep-items-${CSID}"
53	```
54
55	From $ARGUMENTS extract:
56
57	- What: specific failure or anomaly
58	- Where: local / CI / both; which tool or command; which skill or hook if applicable
59	- When: started recently (after change) or always broken; intermittent or consistent
60
61	Unsupported flag check — after all supported flags extracted (`--fast`, `--keep`), scan `$ARGUMENTS` for remaining `--<token>` tokens. If found: print `! Unknown flag(s): \`--<token>\`. Supported: \`--fast\`, \`--keep\`.` then invoke `AskUserQuestion` — (a) Abort (stop, re-invoke with correct flags) · (b) Continue ignoring (skip unknown flags, proceed). On Abort: stop.
62
63	## Step 2: Gather signals
64
65	Init run directory unconditionally at start of Step 2 — `$INVESTIGATE_RUN` must be set even when Step 4 skipped (`--fast` path), so Step 6's read of `$INVESTIGATE_RUN/*-review.md` does not expand to `/codex-review.md` or unset reference. Step 4 creates review files only when adversarial review runs; Step 6 must guard reads with `[ -f <path> ]`.
66
67	```bash
68	export CSID="${CLAUDE_CODE_SESSION_ID:-$PPID}"
69	# timeout: 5000
70	INVESTIGATE_RUN=".temp/investigate/$(date -u +%Y-%m-%dT%H-%M-%SZ)"
71	mkdir -p "$INVESTIGATE_RUN"
72	echo "$INVESTIGATE_RUN" > "${TMPDIR:-/tmp}/investi

People who liked this also viewed

6 picks

Type

investigate

23stars3forks

View on GitHub

How to install investigate?

borda/ai-rig/investigate

$npx -y skills add Borda/AI-Rig --skill investigate

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	<objective>
2
3	Diagnose unknown failures: broken local setup, environment mismatch, tool misbehavior, hook problems, CI vs local divergence, permission errors, runtime anomalies. Gather signals broadly, eliminate hypotheses systematically, report confirmed root cause + recommended next skill. No fixes — diagnosis only.
4	NOT for: known Python test failures with traceback (use `/develop:debug` (requires `develop` plugin)); `.claude/` config quality sweep (use `/foundry:audit`).
5
6	</objective>
7
8	<inputs>
9
10	- $ARGUMENTS: required — symptom, question, or failing command, e.g.:
11	- `"hooks not firing on Save"`
12	- `"codex:codex-rescue agent exits 127 on this machine"`
13	- `"/calibrate times out every run"`
14	- `"CI fails but passes locally"`
15	- `"uv run pytest can't find conftest.py"`
16
17	- `--fast`: optional flag — skip Step 4 adversarial Codex review; use when speed matters more than thoroughness or Codex unavailable.
18
19	If $ARGUMENTS empty or too vague, use AskUserQuestion: "What exactly is failing or behaving unexpectedly? Include the command and any error output you can share."
20
21	</inputs>
22
23	<compaction>
24	Key boundary 1: end of Step 2 (signals.md written to run-dir), before Step 3 rank hypotheses.
25	Key boundary 2: end of Step 3 (hypotheses.md written), refreshed again after each Step 5 probe verdict — so a mid-loop compaction does NOT re-rank or re-probe.
26	Preserve: INVESTIGATE_RUN, symptom.txt, signals.md, hypotheses.md paths; adversarial-review path (codex/challenger) if Step 4 ran; probe ledger (hypothesis → Confirmed/Ruled-out/Inconclusive).
27	Terminal path: end of Step 6 (report + follow-up gate complete).
28	</compaction>
29
30	<workflow>
31
32	Task hygiene: load and follow the protocol below.
33	```bash
34	# loads: compaction-contract.md
35	# audit-skip: resilience-replication
36	_FS=$(python "${CLAUDE_PLUGIN_ROOT:-plugins/cc_foundry}/bin/resolve_shared_path.py" foundry skills/_shared 2>/dev/null \|\| echo "plugins/cc_foundry/skills/_shared") # timeout: 5000
37	cat "$_FS/task-hygiene.md"
38	```
39
40	Task tracking: TaskCreate tasks for Gather, Hypothesise, Probe, Report; mark in_progress/completed as you go.
41
42	## Step 1: Parse symptom and scope
43
44	```bash
45	export CSID="${CLAUDE_CODE_SESSION_ID:-$PPID}"
46	KEEP_ITEMS=""
47	if [[ "$ARGUMENTS" =~ --keep[[:space:]]\"([^\"]+)\" ]]; then
48	KEEP_ITEMS="${BASH_REMATCH[1]}"
49	fi
50	ARGUMENTS=$(echo "$ARGUMENTS" \| sed 's/--keep "[^"]*"//g')
51	rm -f .temp/state/skill-contract.md ${TMPDIR:-/tmp}/investigate-verdicts-${CSID} # clear stale contract + probe ledger (compaction-contract.md §Lifecycle) # timeout: 5000
52	echo "$KEEP_ITEMS" > "${TMPDIR:-/tmp}/investigate-keep-items-${CSID}"
53	```
54
55	From $ARGUMENTS extract:
56
57	- What: specific failure or anomaly
58	- Where: local / CI / both; which tool or command; which skill or hook if applicable
59	- When: started recently (after change) or always broken; intermittent or consistent
60
61	Unsupported flag check — after all supported flags extracted (`--fast`, `--keep`), scan `$ARGUMENTS` for remaining `--<token>` tokens. If found: print `! Unknown flag(s): \`--<token>\`. Supported: \`--fast\`, \`--keep\`.` then invoke `AskUserQuestion` — (a) Abort (stop, re-invoke with correct flags) · (b) Continue ignoring (skip unknown flags, proceed). On Abort: stop.
62
63	## Step 2: Gather signals
64
65	Init run directory unconditionally at start of Step 2 — `$INVESTIGATE_RUN` must be set even when Step 4 skipped (`--fast` path), so Step 6's read of `$INVESTIGATE_RUN/*-review.md` does not expand to `/codex-review.md` or unset reference. Step 4 creates review files only when adversarial review runs; Step 6 must guard reads with `[ -f <path> ]`.
66
67	```bash
68	export CSID="${CLAUDE_CODE_SESSION_ID:-$PPID}"
69	# timeout: 5000
70	INVESTIGATE_RUN=".temp/investigate/$(date -u +%Y-%m-%dT%H-%M-%SZ)"
71	mkdir -p "$INVESTIGATE_RUN"
72	echo "$INVESTIGATE_RUN" > "${TMPDIR:-/tmp}/investi

People who liked this also viewed

6 picks

Type