Rishikesh Ranjan — rishikeshranjan.com

// utility

Archive

/ttm-archive

Archive a completed campaign, finalize state, and update LEARNINGS.md. Use when a campaign lifecycle is fully complete.

manual invocation
allowed-tools:ReadWriteBashGlobAskUserQuestion
SKILL.md ↗

Overview

Archives a completed campaign: it finalizes the campaign state, extracts structured learnings from the campaign artifacts, moves the campaign into the archive directory, and records the lessons in LEARNINGS.md. Use it when a campaign lifecycle is fully complete.

Invoke it with ttm-archive followed by a campaign slug (for example ttm-archive [campaign-slug]). Archiving is the marketing equivalent of merging and deleting a branch: the work is preserved, but the campaign is removed from the active workspace and the ttm-next queue. Archived campaigns still feed ttm-learn.

How it works

Step 0: First-run inline education

Read .taketomarket/CONFIG.md. Parse first_run_seen (object) and inline_education (boolean, default true).

If inline_education is false: skip this step. Else if first_run_seen.ttm-archive is not true, print the explainer below verbatim, then mark this skill as seen:

node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run mark ttm-archive

Use this exact check (bash) to decide whether to print: node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" first-run check ttm-archive --raw — the JSON seen field is true once the explainer has run before.

Explainer for /ttm-archive

/ttm-archive moves a finished campaign out of the active set into .taketomarket/archive/. It freezes STATE.md, preserves the gate results and ship records, and removes the campaign from the /ttm-next queue. Archived campaigns still feed /ttm-learn.

Why it matters: campaigns that should be done but stay marked active pollute scheduling and waste cognitive budget every time you run /ttm-state. Archive is the marketing equivalent of merging and deleting a branch — the work is preserved, but it's out of your active workspace.

(Canonical source: references/inline-education-blurbs.md. Embedded verbatim because workflows do not @-resolve files at runtime.)

Purpose

Archive workflow for /ttm-archive. Validates campaign is shipped or learned (only shipped or learned campaigns can be archived per D-08), extracts structured learnings from campaign artifacts (D-09), moves campaign to ARCHIVE/ directory (D-07), and updates LEARNINGS.md with extracted lessons.

Archive is irreversible (D-10). Once archived, a campaign cannot be un-archived via this command. Cancelled campaigns cannot be archived — they stay in CAMPAIGNS/ as cautionary records (D-08).

Required reading

  • ${CLAUDE_PLUGIN_ROOT}/references/context-loading.md
  • ${CLAUDE_PLUGIN_ROOT}/references/learnings-extraction.md

Constraints

POSITIONING.md is READ-ONLY

Do NOT modify .taketomarket/POSITIONING.md during this workflow.

POSITIONING.md is an architectural invariant. If you detect positioning drift:

  • In verify: use the Escalate option to launch /ttm-positioning-shift
  • In other workflows: flag the issue and recommend running /ttm-positioning-check

Only /ttm-positioning-shift and /ttm-init may modify POSITIONING.md.

Archive is Irreversible

Once archived, a campaign cannot be un-archived via this command. The user must be informed and must confirm before the archive operation executes. This is a destructive operation that moves files to ARCHIVE/ permanently.

Shipped or Learned Validation

Only campaigns with phase = "shipped" or phase = "learned" can be archived. Campaigns in any other phase (including "cancelled") must be rejected with a clear explanation of why and what to do instead.

Process

Text-Mode Detection

Text mode (--text flag): Set TEXT_MODE=true if --text is present in $ARGUMENTS or if the runtime is not Claude Code. When TEXT_MODE is active, replace every AskUserQuestion call with a plain-text numbered list.

Detection:

if echo "$ARGUMENTS" | grep -q -- '--text'; then TEXT_MODE=true; fi

If AskUserQuestion tool is not available in the current runtime, set TEXT_MODE=true.

When TEXT_MODE is active, replace each AskUserQuestion with a plain-text numbered list:

[HEADER]
[QUESTION]
  1. [OPTION_1_LABEL] -- [OPTION_1_DESCRIPTION]
  2. [OPTION_2_LABEL] -- [OPTION_2_DESCRIPTION]
  ...
Type the number of your choice:

Step 1: Load Context

takeToMarket > LOADING CONTEXT

Extract SLUG from $ARGUMENTS (strip --text flag if present):

SLUG=$(echo "$ARGUMENTS" | sed 's/--text//g' | xargs)

If SLUG is empty, error: "Usage: /ttm-archive [campaign-slug]. Provide a campaign slug." Exit.

Step 2: Validate Campaign

takeToMarket > VALIDATING CAMPAIGN

Run:

node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" campaign state "${SLUG}" --raw

Parse output and validate:

  • If exists: false: tell user campaign not found. Exit.
  • If phase equals 'archived': tell user campaign is already archived. Exit.
  • If phase equals 'cancelled': tell user cancelled campaigns cannot be archived per policy. They stay in CAMPAIGNS/ as cautionary records for future reference. Exit.
  • If phase is NOT one of 'shipped' or 'learned': tell user only shipped or learned campaigns can be archived. Display current phase and suggest completing remaining phases: "Campaign '${SLUG}' is in phase '${phase}'. Only shipped or learned campaigns can be archived. Complete the remaining phases first." Exit.

Step 3: Extract Learnings

Skip check: If campaign phase is "learned" (meaning /ttm-learn already ran), skip this step entirely:

takeToMarket > Learnings already extracted via /ttm-learn. Skipping re-extraction.

Proceed directly to Step 4.

takeToMarket > EXTRACTING LEARNINGS

Following the learnings-extraction.md reference guide, scan campaign artifacts.

Read these files from the campaign directory:

  1. .taketomarket/CAMPAIGNS/${SLUG}/STATE.md (full file — frontmatter and body)
  2. .taketomarket/CAMPAIGNS/${SLUG}/MANIFEST.json (if exists — per-asset gate results)
  3. .taketomarket/CAMPAIGNS/${SLUG}/BRIEF.md (if exists — original strategy)
  4. VERIFICATION.md in the campaign directory (gate details)
  5. Any FIX-BRIEF-*.md files (fix attempts and outcomes)
  6. Any REVIEW-FEEDBACK-*.md files (human review feedback)

Extract structured learnings in three categories:

1. What worked

Identify successes from the campaign:

  • First-attempt gate passes (verify.run_count = 1 AND verify.overall_result = pass)
  • High review scores (review.overall_result = approved without fix loops)
  • Assets shipped without fix loops (fix.run_count = 0 or null)
  • Any asset with review_status = "approved" on first submission in MANIFEST.json

For each success, identify the specific pattern or approach that drove it.

2. What didn't work

Identify failures or friction:

  • Gate failures (verify.overall_result = fail on any run)
  • Fix loops (fix.run_count > 0) — check FIX-BRIEF-*.md for root cause
  • Review rejections in REVIEW-FEEDBACK-*.md
  • Multiple verification runs (verify.run_count > 1)

Categorize each using the root-cause taxonomy from learnings-extraction.md:

  • positioning-drift
  • weak-hook
  • wrong-channel
  • bad-timing
  • unverifiable-claim
  • broken-funnel
  • creative-fatigue

3. Campaign-level decisions

Extract key strategic decisions:

  • Positioning anchor chosen (from BRIEF.md)
  • Channel mix rationale
  • Hook strategy used
  • ICP targeting decisions
  • Any deviations accepted during verify

Format each learning as a LEARNINGS.md table row:

| ${today} | ${SLUG} | ${category} | ${lesson} | ${action_taken} |

Where:

  • today = current ISO date (YYYY-MM-DD)
  • category = one of: success, positioning-drift, weak-hook, wrong-channel, bad-timing, unverifiable-claim, broken-funnel, creative-fatigue
  • lesson = one-sentence specific summary of what happened and why
  • action_taken = recommended reference file update, or "none" if no systemic fix

Step 4: Confirm Archive

takeToMarket > CONFIRM ARCHIVE

Display the extracted learnings to the user in a readable format:

## Learnings Extracted from ${SLUG}

### What Worked
${for each success lesson: bullet point with category and lesson text}

### What Didn't Work
${for each failure lesson: bullet point with category and lesson text}

### Campaign-Level Decisions
${for each decision lesson: bullet point with description}

**Total lessons to record:** ${N}

Ask for confirmation using AskUserQuestion (or text-mode numbered list):

Archive campaign '${SLUG}'? This will:
1. Move campaign to CAMPAIGNS/ARCHIVE/${SLUG}/
2. Add ${N} lessons to LEARNINGS.md
3. Set campaign state to 'archived'

This action is irreversible.

Options:
1. Confirm archive -- proceed with archive and learnings extraction
2. Cancel -- keep campaign in current state

If user selects Cancel, exit without making any changes.

Step 5: Execute Archive

takeToMarket > ARCHIVING CAMPAIGN

Run the archive CLI command:

node "${CLAUDE_PLUGIN_ROOT}/bin/ttm-tools.cjs" campaign archive "${SLUG}" --raw

Parse output and verify archived: true.

If the command returns an error:

  • Display the error message to the user
  • Do NOT attempt to manually move files (the CLI handles all filesystem operations)
  • Do NOT update LEARNINGS.md — the archive did not succeed
  • Exit with the error context

The CLI command handles:

  • Moving the campaign directory to .taketomarket/CAMPAIGNS/ARCHIVE/${SLUG}/
  • Updating the campaign state to "archived"
  • Validating the campaign is in "shipped" or "learned" phase before allowing archive

Step 6: Update LEARNINGS.md

takeToMarket > UPDATING LEARNINGS

Important: This step runs only after the archive CLI command succeeds in Step 5. The campaign is now in ARCHIVE/ with phase set to "archived". Writing learnings after archive confirmation prevents data-loss on retry (duplicate rows) if archive were to fail.

Read .taketomarket/LEARNINGS.md.

Duplicate guard: Before inserting rows, scan existing LEARNINGS.md content for any row containing | ${SLUG} | with today's date. If matching rows already exist for this campaign slug and today's date, skip insertion and log: "Learnings for ${SLUG} already present in LEARNINGS.md — skipping duplicate write." This prevents duplicate lesson rows if the workflow is retried after a partial failure.

Find the marker line: <!-- LESSONS BELOW THIS LINE -->

Marker validation (T-07-10 mitigation):

  • Count occurrences of the marker in the file
  • If exactly 1 marker found: insert all extracted lesson rows immediately AFTER the marker line (one row per line, each on its own line)
  • If 0 markers found: fall back to appending rows after the last table row in the Lessons Log table section (look for the last line starting with |)
  • If >1 markers found: fall back to appending rows after the FIRST marker only. Log a warning about duplicate markers.

Write the updated LEARNINGS.md back to disk.

Also update the Summary section at the top of LEARNINGS.md:

  • Increment "Total lessons" count by the number of new rows added
  • Update "Last lesson date" to today's date

Step 7: Confirm Completion

takeToMarket > ARCHIVE COMPLETE

Display completion summary:

## Archive Complete: ${SLUG}

**Campaign:** ${name}
**Archived at:** ${timestamp}
**Lessons extracted:** ${N}

### Lessons Added to LEARNINGS.md
| Date | Campaign | Category | Lesson | Action Taken |
|------|----------|----------|--------|-------------|
${extracted_rows}

Campaign directory moved to `.taketomarket/CAMPAIGNS/ARCHIVE/${SLUG}/`

### Next Steps
- Lessons are now available in LEARNINGS.md for future campaigns
- Run /ttm-health to verify system consistency
- Start a new campaign with /ttm-new-campaign when ready

Success criteria

  • Campaign validated as shipped or learned before archive (D-08)
  • Cancelled campaigns rejected with explanation (D-08)
  • Learnings extracted from campaign artifacts using extraction guide (D-09)
  • Root-cause taxonomy applied to failure categorization (D-09)
  • User confirmed before destructive archive action (D-10)
  • LEARNINGS.md marker validated (exactly 1 occurrence) before append (T-07-10)
  • LEARNINGS.md updated with new lesson rows via marker-based append
  • Campaign directory moved to ARCHIVE/ via CLI command (D-07)
  • Campaign state set to archived
  • Completion summary displayed with all extracted lessons

Output

  • .taketomarket/LEARNINGS.md (updated with extracted lessons)
  • .taketomarket/CAMPAIGNS/ARCHIVE/${SLUG}/STATE.md (phase set to archived, via CLI)

What if this doesn't fit?

Looks like /ttm-archive can't do that yet.

  • Want a new skill? /ttm-request-skill
  • Existing skill needs work? /ttm-improve-skill