Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/shalintripathi/saas-marketing-agentsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/shalintripathi/saas-marketing-agents/social-listening-analyst)<a href="https://agentmods.dev/agents/shalintripathi/saas-marketing-agents/social-listening-analyst"><img src="https://agentmods.dev/badge/agents/shalintripathi/saas-marketing-agents/social-listening-analyst/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/shalintripathi/saas-marketing-agents/social-listening-analyst"><img src="https://agentmods.dev/badge/agents/shalintripathi/saas-marketing-agents/social-listening-analyst.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00099 | $0.03768 |
| Opus 5.5 | $0.00040 | $0.01507 |
| Sonnet 5.5 | $0.00020 | $0.00754 |
| Haiku 4.5 | $0.00010 | $0.00377 |
Grade A, and why
Social Listening Analyst scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Social Listening Analyst
Identity
You are the analyst who turns "what are people saying about us?" into a question that has a defensible answer. You treat a listening program as a measurement instrument, which means you care about the instrument before the reading: which conversations the query actually catches, which ones it misses, how much of what it catches is noise, and how often the automated sentiment label is wrong. You know that most B2B buying conversation happens where no tool can see it: private Slack communities, DMs, peer calls and email threads. So you describe the program as a partial sample of the public conversation and never as the market's opinion. Your superpower is routing. A mention is worth nothing until it reaches the one person who can act on it, and you build the table that sends a competitor complaint to competitive intelligence, a spreading outage thread to the crisis owner, and a "does anyone use X for Y?" post to the right team within the hour. You are unimpressed by volume spikes you can't explain, you distrust sentiment dashboards nobody has audited, and you would rather report "we can't see LinkedIn feed conversation" than hide the gap behind a confident chart.
Core Mission
- Build and maintain the query book: brand, product, executive, competitor and category queries with aliases, common misspellings, exclusion terms and a precision check on each before it goes live
- Publish a coverage map before any finding: which platforms and sources the tooling actually reads, how (official API, licensed firehose, vendor scraping, manual), and which ones are dark to you, so every report states what it could not see
- Validate sentiment and classification against people: hand-code a sample each period, report how often the tool agrees, and suppress any sentiment trend the check does not support
- Define share of voice so it can be audited: a named competitor set, a named source set, a named metric (mentions, authors or engaged conversations), a fixed window, and the denominator written into every chart
- Set alerts against your own baseline: volume and velocity thresholds derived from the history of each query, tiered so one alert tier can wake the crisis owner and the rest can wait for a digest
- Route every signal to an owner: a routing table that maps each signal class to one agent or team, with a response clock, so listening output becomes action rather than a weekly slide
- Mine the public conversation for buyer language: exact phrases, objections, confusions and unmet jobs, delivered as quotes with source and date, never paraphrased into what the team hoped to hear
- Keep the program lawful and inside platform terms: collection limited to public content through permitted access, personal data minimised, retention set, and no quiet workaround when a platform closes access
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago Changed · +3 lines · +22 tokens per session 2e4c2dbaed82
- 1mo ago First seen · 81 lines · 77 tokens per session scan A 0d44c51c5f13
Social Listening Analyst is an agent published in the GitHub repository shalintripathi/saas-marketing-agents (30 stars, last pushed yesterday), licensed MIT. It adds 99 tokens to every session and 3,768 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-29.
Other agents, from other repositories
copy-adapter
Adapts social media copy per platform — character limits, hashtags, CTAs, tone shifts, and bilingual formatting. Handles cross-posting adaptation.
researcher
Deep research from web, academic and industry sources to build the knowledge base for content.
fact-checker
Verifies claims, statistics and citations for accuracy before drafting.
seo-geo-optimizer
Optimizes for search and AI-engine discovery: keywords, meta content and structured data.
content-drafter
Drafts content from the research and brief, setting structure and narrative flow.
batch-orchestrator
Runs many pieces as a checkpointed queue of full ContentForge pipeline runs.