agentleFS
maludbmaludb/learn-claude/skills/progress/SKILL.md

progress

Skill51 starsChanged 5 months agoMITView on GitHub

Progress report for Claude Certified Architect (CCA-F) exam preparation. Use when the user asks "how am I doing", "am I ready for the exam", "show my certification progress", "study report", or wants to know what to study next for the Claude certification.

23 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use report, teach and 1 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • MicrosoftSkill

    report

    Turn an exploration (threads, findings, charts) into a single Markdown report — note, blog post, executive summary, KPI dashboard, slide brief, or multi-section analytical report, with embedded charts.

    Also in 3 projects that use this

  • GoogleSkill

    teach

    Teach the user a new skill or concept, within this workspace.

    Also in 3 projects that use this

  • SentrySkill

    review

    Fixture review skill.

    Also in 4 projects that use this

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 4 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 16 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read progress.md from the data dir.

    line 13
  2. Distinctive:

    As-is: all domains that have any data, at their current accuracy.

    line 26
  3. Distinctive:

    Follow with one summary line: "X untouched / Y in progress / Z strong of 30."

    line 56
  4. Very specific:

    If any history.md line is tagged exam-mode, add a short mock-exam section:

    line 64
  5. Distinctive:

    If there are 2+ mocks, note the trend (improving / flat / declining) across them in date order.

    line 67
  6. Distinctive:

    If there are no exam-mode lines (or no history.md), skip this section.

    line 70
  7. Very specific:

    Every number must come from progress.md / history.md.

    line 76

“Common” is measured: kept by at least 6 in 10 of the 16 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

CCA-F Progress Report: readiness estimate, coverage, and what to study next

Build the report from the user's real data files. Never invent or estimate numbers that are not derivable from them. Keep the whole report to about 1.5 screens and lead with the readiness estimate.

Step 1 — Load

  • Data dir: ${CLAUDE_PLUGIN_DATA}; if unset or empty, use ~/.claude/plugins/data/learn-claude.
  • Read progress.md from the data dir. Parse the Summary line and the Topics table (columns: Topic, Domain, Seen, Asked, Correct, Accuracy, Last studied).
  • If progress.md is missing, or every row has Asked = 0: check history.md for exam-mode lines first — if any exist, skip to Step 6 and report only the mock-exam section (plus a note that there's no drill data yet); otherwise tell the user there is no study data yet, offer to start their first study session with the quiz skill (e.g. "quiz me"), and stop — no report.
  • Read history.md if present. Session lines look like - 2026-07-02 quiz d1-01: d1-01-q01 ✓ B, d1-01-q03 ✗ A (correct: C) — one dated line per session, one ✓/✗ entry per question. Mock-exam sessions use the same entry grammar with exam-mode in place of quiz <topic>: - 2026-07-02 exam-mode: d1-01-q02 ✓ B, d4-03-q01 ✗ – (correct: C), ... (a – letter means the question went unanswered; it still counts in the denominator). If the file is missing, still produce the report but skip the trend and mock-exam sections.
  • Skip any table row you cannot parse (wrong column count, non-numeric counts) and any history line that doesn't match the session-line shape; if you skipped anything, add one line at the end of the report noting how many rows/lines were skipped, then move on. Never abort on bad data.
  • Topic titles come from the frontmatter title of ${CLAUDE_PLUGIN_ROOT}/content/topics/<id>-*.md. If a table row's topic has no matching topic file, keep its stats in every calculation, show it as "(no topic file)" in the coverage map, and mention it in the skipped-data note.

Step 2 — Readiness estimate (lead with this)

Domain weights: d1 = 27%, d2 = 18%, d3 = 20%, d4 = 20%, d5 = 15%.

  1. Per-domain accuracy = sum(Correct) / sum(Asked) across that domain's topic rows, counting only rows with Asked > 0. Also record each domain's total Asked.
  2. Projected scaled score = 100 + 900 × Σ over domains of (weight fraction × domain accuracy), where weight fractions are 0.27, 0.18, 0.20, 0.20, 0.15 and accuracy is a fraction (0–1). Round to a whole number.
  3. Any domain with total Asked < 10 is insufficient data. If one or more domains are insufficient, present the projection both ways:
    • As-is: all domains that have any data, at their current accuracy. If some domain has Asked = 0, the as-is variant is not computable at all — present only the excluded-renormalized variant and say why.
    • Excluding thin domains: drop insufficient domains and renormalize, i.e. score = 100 + 900 × Σ(included weight × accuracy) / Σ(included weights). Label which domains were thin and their n.
  4. State the pass line: 720 on the 100–1000 scale. Say explicitly whether the projection(s) fall above or below it, and that this is a rough study signal from practice drills — not a prediction of the real exam result.

Step 3 — Per-domain bars

One row per domain in a code block. Bar = round(accuracy / 10) █ characters padded to 10 with ░; use ·········· for a domain with Asked = 0. Format:

D1 Agentic Architecture   27%  ████████░░  80%  (n=25)
D2 Tool Design & MCP      18%  ··········   –   (n=0)

Flag any domain with accuracy < 70% AND weight ≥ 20% as top priority on its row and call it out in one sentence below the chart.

Step 4 — Coverage map

Classify each of the 30 topics:

  • ○ untouched — Seen = 0
  • ◐ seen — Seen > 0, but accuracy < 80% or Asked < 5
  • ● strong — accuracy ≥ 80% AND Asked ≥ 5

Display compactly grouped by domain, one line per domain: glyph + topic id + a 2–3 word title (from topic-file frontmatter), e.g.:

D1: ● d1-01 Agentic Loops · ◐ d1-02 Orchestration · ○ d1-03 Subagents · ...

Follow with one summary line: "X untouched / Y in progress / Z strong of 30."

Step 5 — What to study next

Recommend 2–3 topics using the same scoring as the quiz skill: need = 3 if never seen, 2 if accuracy < 70%, else 1 / (days since last studied + 1); score = domain weight × need (weights 27/18/20/20/15). Take the top scorers — this naturally puts unseen topics in heavy domains first, then weak topics. Give each a one-line reason with real numbers ("d1-03 — never seen, D1 is 27% of the exam"; "d4-02 — 40% on 5 asked"). Offer to start a session on the top pick via the quiz skill.

Step 6 — Practice vs exam-mode

If any history.md line is tagged exam-mode, add a short mock-exam section:

  • Score each mock from its own line: count of ✓ / total question entries on that line, as a percent, with its date.
  • If there are 2+ mocks, note the trend (improving / flat / declining) across them in date order.
  • State plainly: mock exams are the honest readiness signal; the drill stats above are formative practice (immediate feedback, self-paced) and typically run higher than exam-condition performance.

If there are no exam-mode lines (or no history.md), skip this section; if the user asked "am I ready", suggest a timed mock as the best next check.

Style

  • Lead with the projected score vs the 720 pass line; everything else supports that.
  • Whole report ≈ 1.5 screens: headline, bars, coverage map, 2–3 recommendations, mock section only if it exists.
  • Every number must come from progress.md / history.md. If data is thin, say so — don't pad.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.