Anki Starter Decks¶

Lightweight Anki-importable flashcard exports for each major topic area.

What these are¶

Starter decks with ~10 strong, practical cards per topic. These are intentionally small — they seed your Anki collection and can be expanded.

Format¶

Headerless TSV (tab-separated values), 3 columns:

Column	Description
front	Question / prompt
back	Answer / explanation
tags	Space-separated tags (Anki convention)

How to import into Anki¶

Open Anki → File → Import
Select the .tsv file
Set field separator to Tab
Map fields: Field 1 → Front, Field 2 → Back, Field 3 → Tags
Set deck name (e.g., grokdevops::kubernetes-troubleshooting)
Click Import

Available decks¶

File	Topic
`datacenter.tsv`	Datacenter ops, hardware, RAID, cabling
`modern-cli.tsv`	jq, ripgrep, fzf, fd, bat, modern tools
`observability.tsv`	Prometheus, Grafana, Loki, alerting, SLOs
`kubernetes-troubleshooting.tsv`	CrashLoop, OOM, pending pods, DNS
`kubernetes-advanced-ops.tsv`	Operators, CRDs, etcd, HPA, PDB
`security-ir.tsv`	Incident response, containment, forensics — stale, no longer regenerated (see note below)
`aws-troubleshooting.tsv`	IAM, EC2, S3, EKS, VPC issues
`gcp-troubleshooting.tsv`	IAM, GKE, VPC, Cloud Logging
`azure-troubleshooting.tsv`	AKS, NSG, Azure AD, storage
`networking.tsv`	TCP/IP, DNS, routing, VLANs, TLS
`storage.tsv`	RAID, LVM, NFS, CSI, PV/PVC
`databases.tsv`	Replication, backup, PostgreSQL, Redis
`helm.tsv`	Charts, values, hooks, rollback
`gitops.tsv`	ArgoCD, Flux, drift, progressive delivery
`container-runtime.tsv`	containerd, CRI, cgroups, namespaces
`kafka.tsv`	Topics, partitions, consumer groups, ISR
`elasticsearch.tsv`	Shards, replicas, cluster health, queries

Stale deck: `security-ir.tsv`¶

This file is a leftover from 2026-03-11 and is not refreshed by make export-anki-starters. Its source deck, security-ir.tsv, was merged into security.tsv by commit 1fc5e162 (2026-03-19), but the exporter's DECK_MAP still points at the old stem, so the security-ir deck now matches no source and the exporter prints no source cards found, skipping on every run. The file is kept rather than deleted; repointing DECK_MAP at security.tsv would change which cards are exported, so it is an owner call.

Other files in this directory¶

make export-anki-starters writes only the starter decks above. The exporter's other modes write into this same directory, so you may also see:

File	Produced by
`due-review.tsv`	`--mode due` — cards due for SM-2 review
`weak-topics.tsv`	`--mode weak` — cards from weak topic areas
`interview-cram.tsv`	`--mode cram` — large balanced cram deck
`topic-<topic>.tsv`	`--mode topic --topic <topic>`

These are per-learner and regenerated on demand; use --output-dir to write them somewhere else.

Relationship to canonical assets¶

These exports are derived from the canonical flashcard decks in training/interactive/knowledge/data/cards/ and the quiz bank. They are separate files optimized for Anki's import format.

The canonical decks use a 7-column TSV schema (id, category, difficulty, tags, question, answer, source_path) and are the source of truth. These Anki exports are convenience snapshots.

Regeneration¶

These decks are regenerated deterministically from canonical flashcard decks:

make export-anki-starters
# or directly:
python tools/export_anki_starters.py

Selection criteria: - ~10 cards per topic with balanced difficulty (3 easy, 5 medium, 2 hard) - Prefer cards with longer, more educational answers - Skip generic or too-short questions - Deterministic: same input always produces same output

Useful flags:

python tools/export_anki_starters.py --dry-run              # preview, write nothing
python tools/export_anki_starters.py --list                 # show deck -> source map
python tools/export_anki_starters.py --deck datacenter      # regenerate one deck
python tools/export_anki_starters.py --output-dir /tmp/x    # write elsewhere

--count sizes the due, weak, topic and cram exports (default 100). Starter decks are always ~10 cards each and ignore it.

Expanding decks¶

To add cards, either: 1. Add cards to the canonical deck in training/interactive/knowledge/data/cards/ and re-run make export-anki-starters 2. Edit the .tsv file directly (maintain the 3-column format) 3. Use gd-learn export: the tools/gd-learn system has built-in Anki export support via gd_learn.anki.export_anki()